This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project
to make the world's books discoverable online.
It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject
to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books
are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover.
Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book's long journey from the
publisher to a library and finally to you.
Usage guidelines
Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to
prevent abuse by commercial parties, including placing technical restrictions on automated querying.
We also ask that you:
+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for
personal, non-commercial purposes.
+ Refrain from automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the
use of public domain materials for these purposes and may be able to help.
+ Maintain attribution The Google "watermark" you see on each file is essential for informing people about this project and helping them find
additional materials through Google Book Search. Please do not remove it.
+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner
anywhere in the world. Copyright infringement liability can be quite severe.
About Google Book Search
Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web
at http : //books . google . com/|
^i«
.U:
i^<*
/.
n
Institute for Medical Research
Rockefeller Institute for Medical Research
TTV
»/.^ M.*t )
?^^^^^^
>V*
Boston
Medical Lib b art
8 the fenway
■iMia
1
Digitized by
Google
Digitized by
Google
Digitized by
Google
Digitized by
Google
Digitized by
Google
0"^/
MONOGRA
ON MEDICAL AND ALLIED SUBJECTS
PUBLISHED BY THE
ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH
No. I
ISSUED JUNE 30, 1910
NEW YORK
66th Street and avenue A
1910
Digitized by
Google
I
Digitized by
Google
ON M
. (\^ [^ JAN24J917
STUDIES UPON A TRANSPLANTABLE RAT TUMOR.
Originally Regarded as a Sarcoma; Probably a Teratoma
FROM Which an Adeno-carcinoma Developed.
By SIMON FLEXNER, M.D., and J. W. JOBLING. M.D.
{From the Laboratories of the Rockefeller Institute for Medical Research,
New York,)
Plates I-XVL
By far the greater number of studies of transplantable tumors
have been made upon mice, in which species tumors not infrequently
arise spontaneously. These tumors consist for the most part of
the epithelial neoplasms which take their origin from the mammary
glands, as was originally pointed out by Livingood, and present the
histological appearances of adenoma and adeno-carcinoma, the
dividing line between the two being indefinite. Other types of
tumors have been found very rarely in mice. Livingood described
a sebaceous adenoma ; Ehrlich, a chondroma ; Haaland and Tyzzer,
lympho-sarcomata ; Bash ford, Murray and Haaland, epidermal
cancer; and, more recently, Jobling has met with an instance of
what appeared to be Hodgkins's disease, and two examples of
papillary cystoma of the ovary. In several instances, namely, in
the cases reported by Ehrlich and Apolant, Leo Loeb, Bashford and
Liepman, adenomata have been described which through hyper-
trophy of the stroma and elimination of the epithelial elements have
changed slowly or more quickly into spindle-cell sarcomata. Once
the sarcomatous condition has become established, it has persisted
through many transplantation generations.
The transplantable mouse tumors have been studied with great
profit, and the high degree of success which has attended these
studies and the considerable advance made in uncovering the biolog-
ical conditions underlying tumor growth have resulted in large part
from the happy circumstance that a fair percentage of spontaneous
tumors of mice are transplantable into individuals of the same race
1
Digitized by
Google
2 Studies upon a Transplantable Rat Tumor.
or species, and also from the facts of the small size, rapid matura-
tion and brief life history of the mouse, which favor the making of
observations on a large and sufficient scale to yield data of scientific
value.
The number of tumors of rats, which have been studied in a
manner similar to the tumors of mice, is small, in spite of the fact
that rat tumors had been transplanted successfully before mouse
tumors. Moreover, the variety of rat tumors thus far transplanted
exceeds the variety of transplantable mouse tumors, for while nearly
all the mouse tumors thus far described have been epithelial neo-
plasms derived from the mammary gland of the female mouse, far
more sarcomata have been encountered in rats than epithelial tumors,
and both have sprung from diverse parts of the body.
In 1889 Hanau transferred successfully an epidermal cancer of
the rat through two generations. The only previously successful
experiments had been made by Novinski and by Wehr. In 1876
Novinski implanted a cancer of the nose of a dog into two others,
in one of which a lymphatic gland metastasis appeared; and in 1889
Wehr inoculated successfully the so-called carcinoma of the penis
of the dog. As we now know, the latter tumor is not a cancer, but
probably a round cell sarcoma.
The instances in which adenomatous tumors of the mammary
gland of the rat of a transplantable nature have been recorded are
very few, the chief one consisting of the instance reported by
Michaelis and Lewin. It is noteworthy that this tumor lost, after
a small number of passages, its adenomatous and took on a sarco-
matous character. We wish to add another word to the subject
of the discrepancy in the histological structure and place of origin
of mouse and rat tumors. The records of the spontaneous tumors
of animals are very imperfect and take account of little more than
the transplantable examples. We know from our own experience
that simple adenomatous and fibromatous tumors arise not so very
infrequently from the mammary gland of the female rat, although
they have not up to the present time been transplanted successfully.
We have records of three such examples of benign fibro-adenomatous
neoplasms, as well as of a lipoma of the subcutaneous tissues.
We have brought together in brief form the recorded instances
Digitized by
Google
Simon Flexner and J. W. Jobling. 3
of transplantable rat tumors. As stated, the first report was by
Hanau (1889), who obsetved three instances of epithelial cancer
among some white rats raised in the laboratory. The cancer affected
the skin of the external genitals. The transplanted tumor arose in
the region of the vulva, and produced inguinal and axillary gland
metastases. Portions of the latter were used for transplantation,
and the secondary nodules showed the histological structure of the
original tumor. In one animal in which the inoculation was into
the scrotum, there were multiple peritoneal nodules, and masses
within the diaphragm, and one mass in the posterior mediastinum.
In 1890 von Eiselsberg reported an instance of a spindle-cell sar-
coma of a mixed gray and white rat, which arose in the periosteum
of the scapula. It was transplanted successfully into the peritoneal
cavity of another rat.
In 1892 Firket reported an example of spindle-cell sarcoma of
the peritoneal cavity of a white rat. A series of successful trans-
plantations was made, and the original histological structure of the
tumor was always reproduced. An effort was made to implant the
tumor upon guinea-pigs, but unsuccessfully.
In 1898 Velich transplanted a sub-periosteal spindle-cell sarcoma
of the thigh of the white rat through several generations. One
series of rats was inoculated into suppurating sinuses without result.
Another was inoculated with blood from the field of incision of the
tumor, also without result. In one rat the grafted tumor pene-
trated into the depth of the tissues, compressed the spinal cord, and
produced motor paralysis. Two guinea-pigs were inoculated with-
out result. Rats inoculated with fluid derived from the tumor, but
free from visible particles, yielded no tumors. Suspension of tumor
particles in salt solution before injection into the peritoneal cavity
did not prevent tumor formation. From the intra-peritoneal masses
subcutaneous tumors were developed in other rats. A second graft
was successfully implanted in a rat in which a growing tumor
already existed. Pieces of tumor kept sterile and outside the body
for twenty-four hours and for three days before inoculation pro-
duced tumors, but after four days none developed. A progressive
decrease in the energy of tumor growth from the first to the ninth
generation was observed, and no further transplantations were
Digitized by
Google
4 Studies upon a Transplantable Rat Tumor.
obtained after the ninth generation. An instance of spontaneous
inoculation of the jaw of a rat was observed in an animal which
gnawed the tumor of a rat kept in the same cage. By wounding a
rat at the back of one of the canine teeth and allowing it to gnaw
a tumor, a nodule was made to develop in the wound, and since
the tumor gnawed was situated in the subcutaneous tissue of the
same rat, the result is an illustration of the development of a second
tumor in an animal possessing a growing tumor.
Leo Loeb in 190 1 transplanted through many generations a cystic
sarcoma of the thyroid gland of the white rat, which was inoculable
under the skin and in the abdominal cavity and by means of the
cystic fluid derived from the tumors. The fat of the implanted
fragments was studied. Of the grafts many of the cells degen-
erated, but a nimiber retained a normal appearance and mitoses
were observed in the periphery as early as fifteen or twenty days
after implantation. Of these some probably belonged to the tumor
cells. In from five to eight days a capsule of connective tissue had
surrounded the graft. The center of the graft at this time showed
shrunken cells with pyknotic nuclei and blood vessels with desqua-
mated epithelium, but not all the cells were necrotic. Mitoses were
visible as late as the twelfth to the fifteenth day, at about which time
the tumor cells invaded the capsule. In order to ascertain the
demarcation between host and tumor cells the grafts were enclosed
in gauze sacs, and in spite of suppuration about many of the sacs,
a part of the grafts developed, but the result was not definitive. It
was noted that when the graft became infected its growth was
retarded. After arrest of growth, excision of a part sometimes led
to renewal of development, and replantation of the excised part to
another part of the body was sometimes followed by growth of the
new graft. In some cases in which the excision failed to awaken
growth in the original tumor the graft derived from it, implanted
in another animal, yielded a tumor. Local metastases were pro-
duced, but systemic metastases did not develop even when the tumor
invaded a blood vessel. The examples of contact inoculation were
most pronounced from intra-peritoneal implantations, and the sepa-
rate nodules even penetrated the diaphragm and projected into the
pleural cavity. The histology remained fixed during fifteen months
Digitized by
Google
Simon Flexner and J. W. Johling. 5
of transplantation. Attempts to implant the tumor into guinea-
pigs, hens and mice failed. In guinea-pigs, at first mitotic division
of the tumor cells occurred, but the grafts did not increase in size
and the emigration of leucocytes was more abundant than about the
grafts in rats. The opinion is expressed that in the production of
tumors by grafting the new tumor is derived from the cells of the
graft and not from the cells of the host; and Loeb, therefore, views
these tumor cells as resembling germ cells in their capacity for
unlimited reproduction.
In 1902 Loeb described several more rat tumors. One was an
adeno-carcinoma originating apparently in the pancreas and not
transplantable. Another depended from the neck, originated in the
thyroid gland and was divided unequally by a fibrous band into
halves, one of which was an adeno-carcinoma and the other a cystic
sarcoma. A single sarcomatous metastasis existed in the lung.
The sarcoma was transplantable. A third rat showed an adenoma
of the mammary gland. It was excised and fragments were im-
planted beneath the skin of the same animal and of other rats. A
month later the fragment implanted in the original rat had increased
to eight times its former size, when the animal gave birth to young,
after which in a remarkably short time the transplanted tumor
shrank in size. No growth took place in the other rats inoculated.
A fourth tumor was a cystic sarcoma of the thyroid gland, which
had produced local and inguinal lymphatic gland metastases. Por-
tions of this tumor kept on ice for five days grew on transplantation,
while other portions kept in the thermostat or at variable room tem-
perature failed to grow. Loeb concluded, erroneously, as we now
know, from these experiments, that two agencies are required to
cause growth of the grafts — the tumor cells and a tumor-producing
factor. The latter he believed did not lose power on exposure to
temperatures of 2° to 4° C, while he believed it less certain that
the sarcoma cells remain alive at those temperatures. He found
that injections of tumor pulp suspended in salt solution and of
filtrates through paper, asbestos and unglazed porcelain gave nega-
tive results. The tumor-producing factor was conceived to be ren-
dered inactive by salt solution. Loeb considered that two factors
were concerned in immunity to tumor grafts, one being a state of
Digitized by
Google
6 Studies upon a Transplantable Bat Tumor.
resistance to the transplanted tumor cells and the other to the tumor-
producing agency.
In 1903 Loeb described in a paper dealing with the endemic occur-
rence of tumors in animals still another example of cystic sarcoma
of the thyroid gland of the rat.
Gay lord and Clowes in 1907 described three rats which developed
tumors while being kept in cages previously employed by Loeb in
his experiments with a transplantable sarcoma of the thyroid gland.
In 1904 a rat was discovered showing a fibro-sarcoma of the abdom-
inal wall and axillary region, and in 1905 two rats were found
showing respectively fibro-sarcoma of the abdominal wall and
spindle-cell sarcoma of the thyroid gland. The last tumor was
transplantable.
In 1907 Jensen reported an instance of transplantable spindle-cell
sarcoma of two gray-brown rats which had previously — and before
the tumors developed — been inoculated with an acid-fast bacillus
derived from a pseudotuberculous disease of cattle. Rat i showed
nodules of varying size in the peritoneal cavity, and small nodules
in the liver and lungs. The growth was infiltrative. Rat 2 showed
small nodules in the lungs and none elsewhere. From Rat i inocu-
lations were made into wild rats and rats obtained from Berlin, but
without results, and into several brown and white rats of the labo-
ratory stock, some of which developed the tumor. Later certain
Berlin and London rats were inoculated with occasional success.
Gray rats, however, were found refractory. The tumor from Rat
2 could also be transplanted to the laboratory stock, but not to the
rats of foreign origin.^
In 1907 Michaelis and Lewin published their studies of a trans-
plantable carcinoma of the rat which in later generations became
converted into a spindle-cell sarcoma. The original tumor devel-
oped in the mammary gland. The first transplantations yielded
about 50 per cent, of tumors and produced lung metastases. Some
of the tumors underwent complete retrogression. It was noted that
young rats were more subject to the implantations than old ones,
and that the white, gray, mixed and black rats were all susceptible
* Professor Jensen kindly sent us living rats possessing tumors, which we
have propagated successfully through many generations in American rats.
Digitized by
Google
Simon Flexner and J. W. Jobling. 7
by subcutaneous and by intra-peritoneal inoculation. Even before
the conversion of the tumor into sarcoma it had been noted that the
microscopical characters of the growth varied considerably and
sometimes presented appearances of pure adenoma ; at other times,
of alveolar and solid carcinoma ; and at still other times, on account
of the rich development of the stroma, of sarcoma. Moreover,
there sometimes developed cystic spaces carrying papillary out-
growths, and the individual cells varied so greatly in respect to size,
quality of protoplasm and nature of nucleus as to excite wonder
that they could all arise from a single tirnior. One tumor of the
third generation is stated to have exhibited in places the appearances
of typical cancroid. The overgrowth of stroma which eventually
gave rise to a sarcoma began in the fifth generation and in the
seventh generation the sarcoma had been perfected. By choosing
the most rapidly growing strains, and by double inoculations at
intervals of a few days, the number of successful implantations was
made to reach lOO per cent. Rats with growing tumors could be
successfully implanted with a second graft. Rats which had recov-
ered spontaneously from growing tumors were resistant to the
several types of this tumor, namely, the adenoma, epithelioma and
sarcoma, and to Jensen's sarcoma. The percentage of spontaneous
recoveries was about ten. The blood of the rat gave a considerable
degree of immunity and a reciprocal immunity was produced between
mice and rats by inoculating with the rat tumor, on the one hand,
and mouse carcinoma, on the other.^
Apolant in 1908 reported very briefly upon twenty- four rat tumors
which had been collected by Ehrlich's laboratory. They consisted
of one fibro-lipoma, two fibromata, three adeno-fibromata, thirteen
sarcomata, and five tumors which were designated mixed tumors
of the " Flexner type."* He speaks also of two forms of rat sar-
*Dr. Lewin kindly sent us specimens of living rats with tumors of the sar-
comatous variety, which we have propagated successfully in American rats
through a number of generations.
'It is perhaps questionable, in view of the transformations through which
the rat tumor we are describing passed, whether Apolant is correct in speaking
of the tumor as a mixed one, and in giving it such a specific cognomen. If it is
true that the tumor we are describing represents merely a type of mixed tumor,
which is not uncommon in the rat, it would be quite significant, we think, par-
Digitized by
Google
8 Studies upon a Transplantable Rat Tumor.
comata according to their microscopical structure. One of these
presented the characters of an ordinary spindle-cell sarcoma, whereas
the other possessed the characters of a round cell sarcoma with
lymph-adenoid structure. The latter type was not homogeneous,
but showed in some parts a cellular structure resembling that of the
lymphatic glands, in which there was a reticulated tissue composed
of spindle and branching cells, in the meshes of which lay the round
cells, and in other parts a looser tissue composed essentially of the
reticulated tissue. Moreover, in still other places the structure
approached the type of the spindle-cell sarcoma. It is stated that
this tumor possessed from the beginning a high degree of energy
of proliferation, and yielded a high percentage of successful inocu-
lations. In the course of the implantations of this tumor it exhib-
ited more and more the character of a spindle-cell sarcoma, until
finally it became a pure tumor of this type. The reverse process
was observed in still another tumor, which, at the beginning, pre-
sented the usual characters of the spindle-cell sarcoma of rats, and
gradually became transformed into a tumor presenting lymph-
adenoid structure.
HISTORY OF THE TUMOR.
A full grown, white, male rat of the laboratory stock died spon-
taneously on January 5, 1906. The autopsy performed soon after
death revealed a tumor the size of a walnut attached to the left
seminal vesicle and projecting into the abdominal cavity. There
were no adhesions to other viscera. The tumor was approximately
spherical in shape and was covered with peritoneum. The surface
was roughened by irregular granulations. The consistence was
firm and the mass cut with difficulty. There were no visible metas-
tases. The cut surface of the tumor was not uniform; it was in
general dead white in color, but certain areas were more opaque
and of a yellowish tinge.
ticularly in view of the manner in which the highly organized type of tumor has
been developed in our example. On the other hand, we are inclined to believe
that the facts to be presented concerning the structure of our tumor with
reference to the original moiety of adeno-carcinomatous structure will suffice
to put it in a place quite by itself.
Digitized by
Google
Simon Flexner and J. W. Jobling, 9
A small piece of the tumor was cautiously removed, cut into bits
with sterile scissors, suspended in sterile normal salt solution and
injected through a needle with a large bore into the abdominal
cavity and beneath the skin of three white rats, the histories of
which will be given later.
The remainder of the tumor was preserved for histological exami-
nation. In the course of the studies of the tumor carried out during
the past few years the original mass has been subjected to repeated
microscopical examinations.
In our first publications* regarding this tumor we described it as
a polymorphous sarcoma. As the following description will show,
the original tumor cannot really be brought into harmony with the
morphology of any of the established types of sarcoma. It would
appear certain now that the original tumor should not have been
regarded as a single type. In the course of transplantation it has
passed through several forms of structure corresponding with sar-
coma and simple and adeno-carcinoma, and has come to be estab-
lished in the last form. When the carcinomatous forms made their
appearance the changes in the structure of the tumor were so
remarkable, and apparently so abrupt, that the origin of the new
elements from the previously perceived ones in the original tumor
seemed exceedingly doubtful. Hence the original mass, which had
been preserved, was subjected to reexamination by sections made
from all parts of it. The result was the discovery of certain struc-
tures of an epithelial character that pointed to the existence of an
adenomatous moiety in the origfinal tumor. These epithelial ele-
ments consisted in part of glandular structures of a normal type,
and in part of such atypical formations as are seen in adeno-
carcinoma. Certain of these glandular bodies of normal appear-
ance seemed indeed to be participating directly in the production of
* Flexner and Jobling, Infiltrating and Metastasizing Sarcoma of the Rat,
Jour, of the American Med. Assoc, 1907, xlviii, 420; Infiltrierendes und meta-
stasenbildendes Sarcom der Ratte, Cent. f. allg. Path., 1907, xviii, p. 257;
Remarks on and Exhibition of Specimens of a Metastasizing Sarcoma of the
Rat, Proc. of the Soc. for Exper. Biol, and Med., 1906-07, iv, 12 ; On Secondary
Transplantation of a Sarcoma of the Rat, ibid., 44.; On the Promoting Influence
of Heated Tumor Emulsions on Tumor Growth, ibid., 156; Restraint and Pro-
motion of Tumor Growth, idem, 1907-08, v, 16; Metaplasia and Metastasis of a
Rat Tumor, ibid., 52; Further Notes on a Rat Tumor, ibid., 90.
Digitized by
Google
10 studies upon a Transplantable Bat Tumor.
the atypical formations indicative of carcinoma, thus indicating pos-
sibly the origin of the carcinomatous elements from several dis-
tinct foci.
It now appeared, therefore, as if the original growth within the
seminal vesicle first consisted of a morphologically simple tumor
made up of one kind of tissue, that by invasion of the inner or
epithelial wall of the vesicle and by inclusion of certain epithelial
structures became, as it were, infected with these epithelial cells,
which upon removal from their normal relations and control assumed
malignant properties. According to this view the epithelial moiety
was inconspicuously present in the part of the tumor that was origi-
nally transplanted, and it gradually divested itself in the course of
the transplantation of the prevailing simple tissue until it became
itself dominant.
It may be questioned whether the part of the tumor originally
described did not possess a greater complexity of structure than
could properly be attributed to a so-called polymorphous sarcoma.
Indeed, it was pointed out originally that the relation of cells and
connective tissue within the original tumor was such as to permit
its being described as imperfectly alveolar. The sections exhibited
a connective tissue framework, varying in density, containing spindle
cells and supporting other cells far richer in protoplasm. The
latter cells occupied spaces of variable size and form, so that some-
times a single cell and sometimes several cells were thus enclosed.
This relation of large cells to stroma was particularly well shown
in sections stained by Mallory's aniline blue and f uchsin method.
The complexity of the original mass does not end here, for the
inclusions at one point were such as to arouse the feeling that the
tumor was, in f act^ originally a teratoma or embryoma, and the sub-
sequent transformations, first into a tumor resembling polymor-
phous sarcoma, and next into an adeno-carcinoma, represented
changes of secondary nature.
HISTOLOGY OF THE ORIGINAL TUMOR.
Under the low power of the microscope the sections showed con-
siderable diversity. Tissue taken from the free part of the mass,
away from its attachment to the seminal vesicle, showed irregular
Digitized by
Google
Simon Flexner and J. W. Jobling, 11
bands and whorls of cells. Two principal types of cells were distin-
guishable: (a) polymorphous cells rich in protoplasm, with single
vesicular nuclei or with multiple or large polymorphous nuclei ; and
(6) elongated and spindle-shaped cells possessing solid nuclei and
closely associated with the connective tissue framework. The ar-
rangement was such that the cells rich in protoplasm lay within
areas composed of spindle cells and connective tissue fibrils. Blood
vessels were fairly numerous and polymorphonuclear leucocytes
richly infiltrated the tissue. Under the medium power of the micro-
scope the distinctions in the two types of cells became more evident.
The large polymorphous cells made up usually by far the larger part
of the tissue, and often were so fused together as to suggest syncitial
formation. In addition, there were present a number of multi-
nucleated giant cells and cells with lobed gigantic nuclei distinct
from the fused small cells. Areas existed containing a preponder-
ance of the spindle cells, but they never occurred strictly indepen-
dently of the larger cells. On the other hand, the spindle cells and
fibrillar connective tissue tended to increase and diminish together.
The collections of poly nuclear leucocytes were larger in the degen-
erated foci, but they were not limited to the larger areas of degen-
eration. The quantity of connective tissue in the tumor was likely
to be underestimated from sections stained in hematoxylin and
eosin (Plate I, Fig. i). The sections stained by the Mallory
method showed the interstitial tissue to be much more abundant than
was suspected, and the polymorphous cells to be surrounded by it.
Sections prepared from those parts of the original mass which
were attached to the seminal vesicle exhibited besides what has been
described still other structures. In the first place, portions of the
wall and epithelial membrane of normal appearance were included.
Moreover, there occurred glandular structures which exhibited evi-
dences of proliferation. These later inclusions were of tubular
nature, although the lumen of the tube was likely to be fUled with
projecting outgrowths of the walls. These formations tended to
be surrounded by structureless proper membranes, but sometimes
the latter were imperfect, so that groups of epithelial cells passed
from the tubule into the adjacent tissue. Where this proliferation
was going on the tubule had lost its regular and characteristic struc-
Digitized by
Google
12 Studies upon a Transplantable Rat Tumor.
ture, and there developed small alveoli, consisting of polymorphous
cells surrounded by spindle-cells and connective tissue, such as com-
posed the mass of the tumor (Plate I, Fig. 2). In another part
large tubules lay in hyalin connective tissue, which separated them
from the usual tumor tissue, so that the formation of definite alveoli
enclosing epithelial cells derived from the tubule was readily made
out (Plate I, Fig. 3).
There was also discovered still another example of carcinomatous
formation of striking and unmistakable character, which originated
close to the epithelial wall of the vesicle. Sections through a part of
the original growth, at a point showing to the naked eye small cavi-
ties and under the microscope cystic dilations of the inner glandular
cavities of the vesicle, exhibited also irregular spaces lined with
cubical epithelium and containing lumena in which there were des-
quamated cells and polynuclear leucocytes. These carcinomatous
acini partially surrounded the cyst-like spaces which were included
in the thickened fibrous wall of the vesicle supporting the tumor
(Plate II, Fig. 4). These carcinomatous acini occurred also
among the tumor cells proper, as well as in the fibrous stroma. In
certain places spindle cells and fibrillated connective tissue en-
croached on the acini, compressing them and bringing about degen-
eration of the cubical cells, and the acini had in places been trans-
formed into solid alveoli and thus made to resemble the cell groups
shown in Fig. i. Sections stained by Mallory's method showed the
stroma to be abundant and to bear that relation to the acini and
alveoli that is seen in a glandular organ.
We come now to a structure discovered within the mass, the
explanation of which is impossible excepting on the basis of a devel-
opmental error. This structure consisted of a large space or tube,
the walls of which were more or less corrugated and branching and
covered with a layer several cells deep of high epithelium. Owing
to the branching nature of the corrugations a certain number of
elongated glandular tubular cavities appeared in the sections to be
contained within the wall. The supporting connective tissue was
richly infiltrated with small round cells. Adjacent to the large tube
there lay a typical compound serous or mucous gland, and outside
that structure a certain amount of smooth muscle existed (Plate
II, Fig. 5).
Digitized by
Google
Simon Flexner and J. W. JobUng. 13
The considerable variety of epithelial formations present in the
tiimor led us to examine microscopically the internal and external
genital organs, and the seminal vesicles especially, of a number of
white rats. Although there was some lack of agreement between
the glandular structures contained in the tumor and those occurring
normally in the vesicle, yet many of them appeared to be identical.
A closer identification was, on the other hand, not possible with any
other of the glandular parts of the internal genital organs. We
failed entirely to identify the large tube with its associated serous
or mucous gland with any structure which we discovered in the sec-
tions of the normal genital organs. It must, therefore, be con-
cluded to have been an accidental and erroneous embryonic inclusion.
The foregoing description of the histological characters of the
tumor indicates that it was not a simple but a complex formation,
in which all the parts were not equal. This inequality was shown
first in the body of the tumor, in which the polymorphous and
spindle cells were united in a heterogeneous fashion, but it was
shown even before in those parts in which the carcinomatous alveoli
had developed (Plate III, Fig. 6). This absence of homogeneous-
ness has an especial significance as regards the understanding of the
interesting series of transformations through which the tumor passed
in the course of transplantation through several thousand rats in
the past three years.
It is not possible to establish the original tumor as a simple out-
growth from the seminal vesicle, as it was at first believed to be.
As has been detailed, the more searching study of the original mass
revealed epithelial structures differing materially from those of the
seminal vesicles. The question, therefore, arises whether the tumor
should not be regarded as a teratoma. If it is to be so regarded,
then it follows that a part only of the epithelial structures contained
within the tumor had been derived from the seminal vesicle, the
other parts and perhaps the chief part of the epithelial inclusions
having been derived from structures which under normal conditions
would go to form adjacent epithelial organs.
It is also impossible to establish the precise origin of the cancerous
transformation which occurred in parts of the tumor, and it cannot
be stated with assurance that the transformation took place at many
Digitized by
Google
14 Studies upon a Transplantable Rat Tumor.
points, rather than at one point. The obvious interpretation to be
put upon the appearances described in connection with those sections
from which Figs. 2, 3 and 4 have been prepared is to the effect that
they represent points at and structures with which this transforma-
tion is proceeding. We rather tend to the opinion that the condition
represented in Fig. 2 speaks for an active transformation of the
glandular tubules into carcinoma, and yet the appearances are not
wholly conclusive, since the reverse process, namely, that of can-
cerous invasion of the tubules from without, can not be excluded.
TRANSPLANTATION GENERATIONS.
In order that the vagaries of the tumor may be followed, and its
properties established, some account of different transplantation
generations will be given. To make the account as brief as possible,
merely characteristic examples will be chosen. The first trans-
plantation generation turned out luckily, since owing to inexperi-
ence only three rats were inoculated. Small morsels cut with
scissors were suspended in salt solution and injected into the peri-
toneal cavity and beneath the skin of three adult white rats on
January 6, 1906. The first of these was killed on February 8, and
no tumor was found. The two remaining rats developed intra-
peritoneal but no subcutaneous tumors. Rat 2 was killed on March
19, at which time a nodule the size of a small bean was attached to
the abdominal parietes. Rat 3 died on June 26, nearly six months
after inoculation. There was no emaciation, but the skin over the
abdomen was tense and the abdominal cavity contained a tumor
the size of a walnut, attached to the abdominal wall and to the
great omentum. The tumor resembled the original in its gross
appearances, although the central portion was necrotic but firm.
Two smaller pea-sized nodules also occurred in the omentum ; each
kidney contained several metastases up to three millimeters in size ;
the lungs were studded with nodules ranging from a millet seed to a
bean in size; a large nodule at the inferior border of the lungs
united them to the mediastinum and diaphragm and extended into
the pericardium; the inferior part of the pericardial sac was
thickened into a cup-shaped depression for the apex of the heart;
Digitized by
Google
Simon Flexner and J. W. Jailing. 15
the left ventricle was invaded from the pericardium and contained
a nodule the size of a pea extending into the ventricle, while the
visceral epicardium was generally thickened and white. On section
the lungs were found to contain disseminated nodules, and the costal
pleura contained several nodules.
It is interesting that Rat 3, which succumbed to the tumor
developing within the peritoneal cavity, should have shown extensive
metastases, while the original animal, which also showed an intra-
peritoneal growth, was not the subject of metastasis. Microscopical
examination of the partially developed nodule from Rat 2 showed
that the growth was proceeding from the periphery, while the
center was already necrotic. The clearest picture illustrating the
mode of growth was found at the point at which the abdominal
muscle was being invaded. The cells of the outermost layer were
small and round and resembled the usual small cell infiltration
observed at the growing edge of malignant tumors in human beings.
Mingled with them were oval epithelioid cells derived from the
muscle and the interstitial connective tissue, and a moderate number
of polynuclear leucocytes. Adjacent to this layer was a layer of
capillary vessels separated from one another by oval and elongated
cells and a fine fibrillar tissue. This one, which was very cellular,
passed over into the fibrous layer, which was more cellular at the
periphery than at the center. This fibrous strand was not strictly
homogeneous, but included islands of large cells corresponding with
the polymorphous cells, rich in protoplasm, of the original tumor.
In the central fibrous tissue small islands of these cells were under-
going necrosis. There were present a certain number of multi-
nucleated masses of protoplasm, constituting giant cells, or resem-
bling syncytium. Sections stained by Mallory's method brought
out the fact that almost everywhere in the tumor the connective
tissue was so arranged as to form definite spaces enclosing oval
and polyhedral cells with large vesicular nuclei. The tumor was,
therefore, not a homogeneous structure anywhere, but was made up
of two elements, connective tissue and cells independent of but
enclosed in it.
The microscopical appearance of the tumor nodules in Rat 3 will
be described. The large omental nodule showed a cortex merely of
Digitized by
Google
16 IStudies upon a Transplantable Bat Tumor.
living tissue, and center consisting of hyalin and fibrous tissue.
The most peripheral zone was the most cellular, but also contained
much fibrous tissue. The cells consisted chiefly of the large poly-
morphous cells, with vesicular nuclei occurring singly and in groups.
The center, which was hyalin and fibrous, still showed areas of the
polymorphous cells. These cells and the cells forming the capil-
laries and the stroma were quite distinct. The smaller omental
nodules represented developments within the adipose tissue. The
most recent invasion was at the periphery, which consisted of areas
of the polymorphous cells embedded in hyalin connective tissue.
These nodules were inactive and the centers, already hyalin, enclosed
atrophying cells of the large type.
The lung nodules might or might not involve the pleura. When
the intrapulmonary nodules were small they surrounded the
medium-sized blood vessels and consisted of the large cells occupy-
ing the perivascular lymph spaces, from which they spread into the
alveoli. Apparently the alveolar epithelium was lost and the spaces
were entirely filled from the larger tumor cells. Unless the pleura
was invaded there was no increase in the connective tissue, but
invasion of the pleura at once led to a new growth of connective
tissue, which together with the tumor alveoli quickly underwent
hyalin and other forms of necrosis. However, in the hyalin tissue
a certain number of columns of the large tumor cells persisted.
Usually the lower border of the tumor in the pulmonary paren-
chyma preserved its integrity. In addition to the usual cells, multi-
nucleated giant cells occurred in the pulmonary nodules. Sections
carried through the large nodule at the hilum of the lung from
which the pericardium was involved showed appearances similar to
those of the large intraperitoneal mass, but in addition there were
present a remarkable number of mono- and multinucleated giant
cells, some of which were of the megacaryocytic type (Plate III,
Fig. 7). These cells often included mononuclear and polynuclear
leucocytes. Some of the massive giant cells included scores of
smaller cells, probably leucocytes, in process of fragmentation.
The giant cells were irregularly distributed. The great blood
vessels were surrounded by tumor, but were patent. An inter-
mediate stage was present in this part of the lungs, between the
Digitized by
Google
Simon Flexner and J. W. Jobling. 17
early condition in which the alveoli were occupied by tumor cells,
and the later one, in which the tumor had become hyalin and fibrous.
This intermediate stage was characterized by a thickening, through
spindle cells and fibrillated tissue, of the air vesicles, which con-
tained mononuclear cells of the type of the tumor cells, suggesting
when viewed under the low power of the microscope an alveolated
tumor of the epithelial type. Mitotic figures in the large cells were
frequent. The nodules in the parietal pleura were invading the
intercostal muscles on the one side and projecting into the pleural
cavity on the other. They lacked the organoid characters described
and were composed of spindle cells coursing longitudinally and
transversely, and of small round cells next to the muscle, and were
invaded with polynuclear leucocytes.
The masses in the kidneys, the largest of which measured six by
four millimeters, developed in the cortex. These nodules, some of
which were wedge-shaped, were produced by an invasion of the
tubules of the kidney and of the perivascular connective tissue with
the large polyform tumor cells, which had taken the place of the
proper epithelium of the tubules. Among them a few multi-
nucleated giant cells were present. The glomeruli persisted much
longer than the tubular structures and the intertubular connective
tissue was increased by a layer of fine connective tissue fibrils and
spindle cells. Polymorphonuclear leucocytes were distributed
throughout, and accumulated most numerously in minute foci of
necrosis. When the nodules projected beyond the capsule they
consisted chiefly of connective tissue containing a certain number
of large cells occupying definite spaces suggestive of lymphatics.
When these lines of cells were longitudinally distributed, they some-
times extended across the field of the microscope, and when they
were cut obliquely or transversely, they resembled alveoli.
The nodule in the ventricular wall consisted of a solid tumor
composed of a fibrillated stroma, spindle cells and capillary vessels,
in which the large tumor cells were embedded in an irregular
manner. There was absence of alveolation, and the tumor cells
proper formed small groups or occurred singly, and were closely
surrounded by the stroma. Giant cells were also numerous. At
the extending edge of the nodule a zone of small round cells
Digitized by
Google
18 Studies upon a Transplantable Rat Tumor.
appeared, from which capillaries penetrated the muscle fibers.
These fibers were separated and compressed by spindle and small
round cells, a formation disclosing the invasion of the tumor cells
proper, which passed between the muscle fibers in parallel lines.
The cells that appeared to be the most recent consisted of multi-
nucleated masses of protoplasm. What was taken to be the coro-
nary sinus had been invaded by the large cells, which were unpro-
vided with a stroma. At one point the lumen was bridged by the
tumor cells and at other places a fibrinous thrombus had developed,
into which the tumor cells were pushing (Plate IV, Fig. 8). The
general layer of epithelium was thickened and infiltrated with a
mixture of the large tumor cells and spindle cells, and was excess-
ively vascularized. It is patent that this transplantable tumor
possessed a considerable degree of invasive power, while it
exhibited, on the other 'hand, relatively slow growth. Death was
caused apparently not by interfering with the nutrition of the
animal, but rather by mechanical causes, chiefly probably connected
with the extensive lesions of the lungs.
The structure of the tumor, as revealed by the first transplanta-
tion nodules, was sufficiently varied to make classification very
difficult. The tumor was not, apparently, homogeneous in nature,
as are the usual sarcomata, but rather of heterogeneous composition,
such as is found among the epithelial tumors. The point should be
emphasized that all the metastases discovered were such as could
best be produced through the general blood current. There was
no evidence of involvement of the lymphatic glands anywhere, but
that the lymphatic vessels of the viscera could be invaded was indi-
cated by the conditions observed in the lungs and the capsule of the
kidney. Direct infiltration of the serous membranes through im-
plantation occurred in the peritoneal cavity and in the epicardium.
In the course of the different generations considerable variations
in histological structure and metastatic properties, which it is not
necessary to describe in detail, were presented by the tumor. On
the other hand, it is desirable to present the important facts regard-
ing the peculiar behavior of the tumor. , The generations represent
a continuous series, while different strains within a generation,
expressed by letters of the alphabet, do not represent a uniform
Digitized by
Google
Simon Flexner and J. W. Jobling. 19
series, but indicate merely the different sets of inoculations, usually
from different nodules, within a generation. A given generation
will, therefore, contain one or more strains according to the circum-
stances governing the transplantations. These facts are of some
importance, in view of the circumstance that the tumors were often
histologically dissimilar within a given generation, as will be shown.
ABSTRACT OF PROTOCOLS OF THE RATS WITH TUMORS.
In this section there will be presented a selected series of pro-
tocols, briefly abstracted in order to bring out the varying and
changeable properties of the tumor.
Of the second generation, strain a. Rat 7 survived four months and pre-
sented a large ulcerating subcutaneous mass and pulmonary and diaphragmatic
metastases. Rat 8, inoculated at the same time as 7, developed a subcutaneous
tumor which ulcerated later, as a result of which it survived a month longer
and showed only pulmonary metastases. Rat 6, of the same series, survived
two weeks longer than Rat 7, and showed, besides the subcutaneous nodule,
kidney metastases.
Rat 64 of the fourth generation (a) survived five and one-half months. The
tumor which first developed was partially excised for inoculation purposes, but
it recurred and eventually caused death. It invaded the muscles and peritoneum
and projected into the abdominal cavity posteriorly. Metastases were present
in the sternum, cartilage of the ribs, diaphragm, left thoracic wall to a marked
degree, right thoracic wall to a small degree, the anterior mediastinum, the
lungs and left kidney. The slower progress of the tumor was the result of
the excision, since other rats of this series not operated upon succumbed within
three to four months of the inoculation. Other rats showed lung metastases
and one. Rat 66, a metastasis in the auricular appendage. Rat 47 is of special
interest in this connection. It was inoculated into the peritoneal cavity with a
single fragment of tumor and survived about four months. The main tumor,
which developed from the fragment, was situated in the omentum, but sec-
ondary nodules were present elsewhere in the abdominal cavity (Plate XIII,
Photograph i). One of these, measuring i by 2 centimeters, was attached to
the ensiform cartilage, which proved to be a point of predilection for the
development of intra-peritoneal nodules. A second point of predilection, as it
afterwards proved to be, was the space between the liver and the diaphragm,
next to the cardiac extremity of the stomach. In this instance the stomach had
been invaded by the tumor, which had passed to the mucous membrane and
led to a partial digestion of the membrane and production of a cup-shaped
depression. A single pin-point metastasis existed in the lungs. Rat 46 was also
inoculated into the peritoneal cavity. It survived about three months. The
original fragment developed in the pelvis to a tumor measuring 3 by 2 by i^
centimeters, which had invaded the muscle of the abdominal wall. A series of
Digitized by
Google
20 Studies upon a Transplantable Eat Tumor.
secondary nodules developed within the omentum and extended upwards to the
transverse colon and in the gastro-hepatic omentum, reaching the diaphragm and
binding it tp the lesser curvature of the stomach. The cardia of the stomach was
invaded over its entire extent at the diaphragm by the tumor, which passed into
the cavity and replaced the mucous membrane, which became altered to form
a cup-shaped depression (Plate XIV, Photograph 2). Separate small nodules
appeared over the peritoneal covering of the stomach, and the lungs showed a
number of small metastases. Rat 61 survived a subcutaneous inoculation about
three months. The sup'erficial tumor which developed at the lower border of
the thoracic wall measured 3 by 4 centimeters. It penetrated the muscles and
projected into the abdominal cavity, pushing the retro-peritoneal tissue before it.
As a result the right kidney and adrenal gland were forced forwards, the
kidney was extensively invaded with tumor, the right adrenal gland was con-
verted into tumor, and a mass of tumor extended in the gastro-hepatic omentum
towards the lesser curvature of the stomach, over which it spread and into
which it penetrated. An elevated boss-like nodule appeared in the mucous
membrane and was connected with the mass in the wall; while on the opposite
(left) side a larger ulcerated area existed, probably representing an implanta-
tion tumor upon the mucous membrane through contact (Plate V, Fig. 9).
Rat 67 survived five months and was inoculated on two separate occasions. The
first tumor developed in the subcutaneous tissues and ulcerated. The second
fragment was introduced into the lumbar muscles, developed there and ex-
tended upward, invading the sternum and ribs. The left pleural cavity was
quite -filled with a mass which compressed the left lung entirely. At the same
time it grew downwards, displacing the diaphragm inferiorly, and upwards to
the apex of the thoracic cavity, pushing through the ribs and intercostal muscle
and came to lie beneath the superficial fascia. The pericardium and mediasti-
num were overgrown and the right pleura, which was invaded, led to the binding
of the lung to the ribs. Rat 51 survived an intra-abdominal inoculation about
four months. A main mass 5 by 5J^ centimeters developed and secondary small
nodules were produced. The mesentery was invaded, thickened, and caused to
retract, so as to bring the small intestine into a close small coil. The right
kidney was surrounded by a mantel of tumor; the diaphragm, and from it the
pleura, were invaded, and ascites existed. The ascitic fluid measured 25 centi-
meters and contained masses of tumor cells.
As will be observed, in the course of the fourth generation, in which a con-
siderable number of rats was inoculated, a considerable variety of effects were
produced. What is particularly striking is the invasiveness exhibited by the
tumor and its tendency to develop frequently in the region of the diaphragm,
adjacent to the right lobe of the liver and the stomach, and to avoid invading
the former and choosing regularly to invade the latter. In no instance during
this generation was a lymphatic gland metastasis observed.
In the fifth generation some interesting examples were observed. Rat 325
{g) was inoculated into the muscles of the back and survived 25^ months. The
tumor which developed measured 2 by 2 centimeters. It invaded the ribs and
spinal column in the dorsal region, produced paraplegia, and, projecting into the
thoracic and abdominal cavities, displaced forward the liver and kidney. The
Digitized by
Google
Simon Flexner and J, W, Johling. 21
lungs and mediastinum contained nodules. Rat 327 was killed after three
months, at which time there existed besides the subcutaneous nodule and the
lung metastases a mass in the region of the axillary gland which measured 0.4
by 0.4 centimeter. It should be remarked that this was the first example of
apparently a lymphatic gland metastasis. Rat 276 survived 2^ months. A sub-
cutaneous nodule of the right side had penetrated the abdominal cavity, pushed
the kidney and adrenal gland forward, displaced the spinal column to the right,
and given rise to a secondary nodule compressing the ureter, which above the
nodule was dilated to the size of a goose quill, and was associated with a hydro-
nephrosis in which the kidney was enlarged to double its normal size. There
were other nodules in the omentum and serous covering of the spleen and liver,
in the mediastinum, and between the heart and the lungs. Rat 196 (d) sur-
vived three months. The original ulcerated subcutaneous nodule involved the
skin and abdominal muscles and penetrated the lower ribs into the abdominal
cavity. In the subcutaneous tissue above the main tumor was a separate smaller
nodule connected with the original tumor by a white cord. The second nodule
was in turn connected by a similar white cord with a third nodule, situated in
the axillary space and representing probably the much enlarged axillary glands.
The white cords were taken to be lymphatic vessels invaded by tumor. Rat
197 showed, as the result of a second implantation, tumor invasion of the
inguinal lymphatic gland on the side of the inoculation, and other metastatic
nodules. Rat 230 exhibited invasion of the ribs, penetration of the thoracic
cavity and the spinal canal, associated with paraplegia. The infra-clavicular
lymphatic glands on both sides were invaded. The superior lobes of the lungs
were studded with nodules, the caudal lobe was the seat of a diffuse infiltration,
a lymphatic gland above the diaphragm was enlarged by tumor to the siz6
of a pea, and the kidney contained a nodule 2}4 centimeters in size. Rat 231
showed a similar condition to the last, and in addition, a retro-peritoneal and
bronchial gland, as well as the left axillary gland, contained metastases (Plate
XIV, Photograph 3; Plate XV, Photograph 4). Rat 93 (a) survived 3^4 months.
A secondary graft developing subcutaneously involved the axillary gland, which
became enlarged to a size of i by 2 centimeters and from which a growth sur-
rounded the clavicle, penetrated into the pleural cavity, and encircled the great
vessels of the heart. The lungs presented nodules, but the heart did not. Rat
88 survived an intra-muscular inoculation two months. The main tumor forced
its way into the thorax on the left side, pressed against the vertebral column,
which was displaced anteriorly and to the right. The ribs corresponding to the
main portion of the tumor had disappeared, the lower half of the left pleural
cavity was obliterated by the mass, which invaded the diaphragm, penetrated
into the abdominal cavity, grew into the cardia of the stomach, the superior
pole of the kidney and the adjacent adrenal gland and involved the Ijrmphatic
glands at the superior surface of the liver. Rat 418 (h) developed a subcu-
taneous tumor which ulcerated and led to axillary and inguinal Isrmphatic gland
metastases. In the inguinal region a chain of enlarged glands or nodules
extended from the margin of the tumor to the angle of the leg. Pulmonary
and mediastinal metastases occurred. Rat 424 (i) was inoculated intra-peri-
toneally with ascitic fluid from Rat 51. Multiple intra-abdominal nodules de-
Digitized by
Google
22 Studies upon a Transplantable Rat Tumor,
veloped in the omentum at the ensiform cartilage and between the spleen and
the left kidney. A chain of nodules extended from the pelvis to the stomach,
the wall of which at the expanded esophageal part was invaded and the mucosa
ulcerated (Plate XV, Photograph 5). Rat 425 developed an intra-abdominal
tumor growing about the stomach, spleen, left kidney and adrenal gland. A
nodule had passed completely through the spleen, and the left adrenal gland
was lost in the growth. Rat 426 exhibited two large and many small intra-
abdominal tumors. The inoculation was made with ascitic fluid in the manner
of Rat 424. The left kidney was replaced by tumor, except for a thin shell of
renal tissue surrounding a large cyst.
In the fifth generation, therefore, the tumor began to produce lymphatic
gland metastases, and at the same time continued to be deeply invasive of tis-
sues in its locality, to be disseminated by implantation, and to produce, as
before, abundant distant metastases. In the succeeding generations the prop-
erties of the tumor as now developed did not undergo any striking alteration.
A few notes may be made, therefore, with reference to certain points which
appeared from time to time. For example, it was noted in the seventh gen-
eration that seven additional examples of gastric invasion occurred. The
invasion always took place at the dilated esophageal extremity, and from the
tumor growing between the liver and the diaphragm. The kidneys were not
infrequently the seat of metastasis. Sometimes one kidney would be quite
converted into tumor (Rat 597, seventh generation), another example of com-
pression of the ureter, succeeded by hydronephrosis, was noted in the tenth
generation. The pulmonary metastases were much the most common of all.
As a rule, when lung nodules existed, the mediastinum was invaded and usually
the parietal pleura and the intercostal muscles became infected. Sometimes an
entire lung was invaded from the hilum (Rat 1775, tenth generation), and com-
pression of a branch of the pulmonary artery by tumor gave rise to infarc-
tion of a lobe of the lung (Rat 401, seventh generation). The myocardium was
invaded in several additional instances. A mass extended from the lungs into
the wall of the heart (Rat 553, sixth generation). A metastatic nodule appeared
in the left ventricle (Rat 401, seventh generation) and implantation nodules
occurred in both auricles and in the left ventricle (Rat 2392, ninth generation).
The lymphatic metastases continued to appear. They were usually in the
axillary and inguinal glands, and arose from subcutaneous inoculation. How-
ever, other superficial and even the deeper glands were sometimes involved.
The retro-peritoneal glands (Photographs 3 and 4) were several times involved,
and once a chain of these glands was affected, along with the axillary gland
(Rat 1648, eleventh generation). A pelvic lymphatic gland was once involved,
but afterwards suppurated (Rat 3035, fourteenth generation). The bronchial
lymphatic glands, together with the axillary glands (Rat i860, tenth genera-
tion) and once the bronchial, trachaeal and clavicular glands were invaded
jointly (Rat 1617, eleventh generation). The inguinal, axillary and mesenteric
glands were once jointly invaded (Rat 2048, tenth generation). In one instance,
a tumor nodule pressing upon the common bile duct produced jaundice (Rat
1338, ninth generation). In two instances the wall of the thoracic duct showed
small nodules (Rats 1753, tenth, and Rat 2273, eleventh generation).
Digitized by
Google
Simon Flexner and J. W, Jobling, 23
So much has been stated of the tendency of the large tumor masses to set
up metastases that it remains to be added that the tumor could grow to a con-
siderable size, produce marked local destruction of tissue, without at the same
time yielding any discoverable metastases.
The object of presenting such a considerable number of protocols
is to bring out the fact of the malignancy of the tumor under dis-
cussion, and the great variety of effects which it produced according
to the circiunstances of its development. The tumor was found to
recur after what appeared to be complete surgical extirpation, to
grow through the skin, muscle, fascia and bone, and to penetrate
into the circulation, thus setting up secondary or metastatic foci at
a distance. It should be noted that until the fifth generation there
had been no instance of lymphatic gland metastasis, and during that
generation and in subsequent generations this form of local metas-
tasis became common. As will be observed a little later, the acquisi-
tion of the property enabling the tumor to invade the lymphatics
was associated with a significant histological alteration in its struc-
ture. The growth within the abdominal cavity tended to become
multiple, even when they arose from a single inoculated fragment,
a result to be expected in view of the conditions in the peritoneal
cavity which favor the secondary implantation of tumor cells. It
is, however, noteworthy that these implantations were so common
between the diaphragm and the stomach, which structures tended
themselves always to be invaded. Of all the intra-abdominal
organs, the stomach was most frequently penetrated by the tumor,
which tended to appear in the mucous membrane and to set up
ulceration quite as tumors do in man which originate in the mucous
membrane. It is further remarkable that the gastric nodules
always developed in the esophageal segment of the stomach, and
that the mucosa was subject to infection directly through contact
with tumor on the opposite side of the viscus. It is worth con-
sidering, therefore, whether the tumor can be implanted on the
normal mucosa. We have abundant evidence to show that rats
cannot be infected by feeding upon the tumor. It was repeatedly
observed that when fragments of the tumor were inoculated into
the muscles they tended to develop towards the interior, and when
implanted beneath the skin, towards the surface of the body.
Digitized by
Google
24 Studies upon a Transplantable Bat Tumor,
Cartilage and bone offered no impediment to the growth of the
tumor. The deposit by preference of the tumor emboli in certain
organs, such as the lungs and kidneys, is a distinctive feature of the
tumor, as is the tendency of the lung nodules to lead to infection of
the mediastinal tissues and the walls of the thorax, the latter appar-
ently through contact. Metastases were never developed in any of
the internal organs or lymphatic glands except through blood in-
fections, although they might be produced by direct extension from
neighboring parts. It is probable that the nodules in the heart
muscle were all the result of ingrowth from the lungs, and it is
remarkable that with the frequency of the pulmonary metastases the
bronchial lymphatic glands were so rarely affected. The superficial
lymphatic glands were much oftener affected than the deep ones.
Small and medium-sized and ulcerated tumors, in which only a thin
membranous remnant of tissue remained, might produce many
metastases, and large, well developed tumors might produce no
metastases. It should be noted that in one instance a nodule com-
pressed the common bile duct and led to jaundice; in two instances
small nodules were observed in the thoracic duct.
HISTOLOGY OF THE LATER TRANSPLANTATIONS.
It is not our intention, in considering the histology of the trans-
plantation tumors, to describe exhaustively the successive genera-
tions. The effort will be made, however, to present the facts of
the histology in such a manner as to bring out the significant and
fundamental histological changes through which the tumor has
passed, and with which have been correlated certain properties of
metastasis, and to make clear minor fluctuations in structure that
are being slowly eliminated. In another section there will be dis-
cussed the manner of proliferation of the tumor fragments, so that
this point can be passed over now.
The second and third generations of the tumor showed no
remarkable variations of the structures already described (see
page 19). A variation was found according as the tumor was
proliferating or had come to a temporary standstill. According to
one or the other, two quite independent types could be distinguished :
Digitized by
Google
Simon Flexner and J. W. Jobling. 25
(i) a proliferating tissue, producing stroma and columns of
epithelial-like cells (Plate V, Fig. lo; Plate VI, Fig. ii ; low and
high power) ; (2) a stationary tissue composed of capillaries and
spindle cells and few epithelioid cells (Plate VI, Fig. 12). The two
types were grossly so different that had it not been for their com-
mon origin and certain resemblances between the epithelial cells they
could not be considered as related. What should be perhaps
emphasized is the fact that in retrogressing hyalin fibroid tissue
associated with the stationary nodules the most persistent cells
were certain epithelioid elements occupying compressed spaces in
the hyalin mass (Plate VII, Fig. 13).
In the fourth generation (a) the appearances were various. An
omental nodule in Rat 45 was composed wholly of elongated, closely
packed cells, without arrangement into alveoli, and the nodule was
quite solid except where islands of adipose tissue remained free of
tumor. In Rat 49 the pancreas had been invaded and the tumor
had produced obliteration of the splenic vessels, causing necrosis of
the spleen. In respect to histology, this tumor was imperfectly
alveolated or fibrous. The growth in the pancreas was highly
fibrous and included isolated tubules resembling ducts, lined with
superficial epithelium, doubtless the remains of the excretory ducts
of the gland. The walls of the obliterated splenic artery and vein,
which contained a hyalin and leucocytic thrombus, showed inflam-
mation and the passage of the large tumor cells from without into
the lumina (Plate VII, Fig. 14). In other sections the splenic
artery showed an obliterating endarteritis of high degree, the spleen,
a wedge-shaped growth of tumor about the vessels, and the liver,
a superficial invasion attended with necrosis of adjacent hepatic
lobules. In Rat 50 the small intrapulmonary nodules consisted of
the air vesicles, in which the normal lining had been substituted by
the large tumor cells. The pleural nodules were larger and the
epithelial masses within them exhibited an acinous arrangement.
When the pleural mass projected beyond the membrane the number
of living tumor cells was small and it was composed chiefly of
hyalin fibrous tissue, or a homogeneous tissue into which the large
pale degenerating tumor cells with large hypochromatic nuclei were
being fused. The portions of the nodule immediately within the
Digitized by
Google
26 Studies upon a Transplantable Rat Tumor,
pleural membrane and extending superficially into the lung were
composed of large, irregular alveoli containing epithelioid cells,
often enclosing clefts suggesting lumena. Some of these alveoli
had become dilated into small cysts enclosing cell detritus. The
large epithelioid cells were dividing rapidly by mitosis, while no
mitoses were seen in the stroma. The tumor, which was extending
into the parietal pleura and the intercostal muscles, preserved its
alveolar character, although the amount of stroma being produced
was highly excessive. The manner of gastric invasion was char-
acteristic (Plate VIII, Fig. 15). In the first place the growth was
limited to the dilated esophageal segment. The section viewed from
the inner surface of the organ showed a cup-shaped depression over
the central and part of the peripheral portion of the growth, which
rested on necrotic tissue devoid of nuclei. The mucous membrane
was persistent at each side of this crater. The main or central mass
of tumor was highly fibrous and resembled in structure a cicatrizing,
scirrhous carcinoma in man. This dense fibrous portion still
showed a small number relatively of compressed areas of cells,
originally the large epithelioid cell. A sharp but iregular line of
demarcation composed of fragmented leucocytes and bacteria
marked the limits of living and dead tissue. The extreme edges of
the tumified mass showed more tumor tissue proper, less fibrous
tissue, and some remains of the smooth muscular wall of the organ.
Where the mucous membrane was still preser\'^ed, the epithelioid
tumor cells occupied spaces between the connective tissue fibrils that
formed the submucosa and penetrated the papillary layer. As soon
as the papillary layer was invaded, two kinds of pathological changes
took place in the mucosa. The first affected the inner or super-
ficial face of the epithelium; the continuity of the horny layer was
disturbed and collections of fluid in vesicles formed. These vesicles
penetrated only a short distance into the substance of the inter-
papillary epithelial processes, but they projected for a much greater
distance into the cavity of the organ. The contents of the vesicles
were homogeneous and contained leucocytes and fragments of
keratinized epithelium. The epithelial layer below was much dis-
turbed. The second change consisted of a deepening of the inter-
papillary processes, which also became narrowed and penetrated
Digitized by
Google
Simon Flexner and J, W. Johling. 27
more deeply into the tumor-infiltrated sub-mucosa. Moreover,
they had divided irregularly into digits. The epithelial cells com-
posing them had become paler and richer in protoplasm and were
actively dividing by mitosis. As the edge of the crater was
approached the dilatation was most marked and the horny layer
imperfectly formed and rapidly disintegrating. In other words,
the epithelial layer of the esophagus, at the point of beginning
invasion of the tumor, was caused to undergo atypical proliferation.
Rat 6i, which also showed invasion of the gastric wall, was interest-
ing in virtue of the fact that it brought out the difference in vulner-
ability of the esophageal and glandular portions of the organ. The
growth was mainly confined in the muscular wall and the peritoneal
coat, both of which were lost in the mass. The mucous membrane
of the esophageal segment was invaded and ulcerated, the invasion
ceasing a fraction of a millimeter before reaching the glandular
segment. At the point of cessation in the esophageal territory the
muscularis mucosae formed the boundary line, although the tumor
cells had penetrated between it and the muscular tunic, which had
now become widely separated. In the normal organ the demarca-
tion between the two anatomical parts — ^the esophageal and
glandular portions of the organ — is distinct and indicated by a
finger-like projection of sub-mucous connective tissue and smooth
muscle fibers, as well as by the change in epithelium. The
squamous epithelium ends on the glandular side of the projection,
although the esophageal layer has lost its papillary character and
become thinner before it disappears. The limits of the tumor were
shown also on the outer surface of the stomach, since the peri-
toneal and muscular tunics, corresponding with the glandular part
of the organ, showed no invasion, and the omental adhesions were
limited wholly to the mass in the esophageal region.
The kidney nodules were developed chiefly in the cortex, the
proper struture of which was replaced by the tumor cells. The
latter invaded the tubules and took the place of the normal epithe-
lium, from which they were distinguished by their greater density
of protoplasm and greater richness in chromatin of the nuclei. The
tumor cells formed a lining similar to the old, but sometimes a double
one. The inter-tubular tissue was increased ; it had become homo-
Digitized by
Google
28 Studies upon a Transplantable Bat Tumor.
geneous and rarely showed the presence of the large tumor cells,
but the compression and obliteration of the tubules, the epithelium
of which was replaced by tumor cells, through the new connective
tissue growth, sometimes produced appearances resembling an extra-
tubular invasion. A certain amount of glomerular obliteration had
taken place through hyalin thrombosis. The central mass of the
tumor, except small islands immediately surrounding blood vessels,
was entirely necrotic, and the walls of the vessels had been invaded
and the endothelium had sometimes been displaced inwards by the
tumor cells, although the lumena were still patent.
The subcutaneous tumors became attached very late to the skin,
because of the great resistance to invasion exercised by the corium
(Fig. lo). Examples show that the growth in the corium tends
not to be continuous, but to be initiated apparently by large epithe-
lioid cells, which are carried to a distance apparently in the tissue
spaces, where they proliferate and start the production of a young
stroma. The stroma of the tumor proper is sometimes more sub-
ject to degeneration than the epithelioid cells. An example of
this condition was presented by Rat 53, in which the stroma was
necrotic and freely invaded with leucocytes undergoing fragmenta-
tion, while the epithelial strands retained their form and appearance,
except for the fact that the protoplasm was more strongly baso-
philic than usual. The numerous blood vessels were dilated and
many contained hyalin or agglutinative red corpuscular thrombi.
From the description of the foregoing examples it would appear
that the tumor had acquired in the fourth generation the unmistak-
able characters of a carcinoma, yet in point of fact the carcinoma-
tous type was not really established in this generation, since the
tendency for the epithelioid cells to be in close relationship with the
stroma had not been entirely lost. Rat 57 illustrates this point.
The tumor when about one centimeter in diameter was in part
excised for purposes of transplantation. The cortical zone, which was
employed, showed under the microscope elongated bundles of highly
cellular connective tissue enclosing small groups of epithelioid cells.
The young tissue consisted of spindle cells which were particu-
larly massed at the periphery, while the central part was more fibrous
or even hyalin. While the epithelioid and connective tissue elements
Digitized by
Google
Simon Flexner and J. W, Jobling. 29
were intermingled at the periphery, the former were well defined
in the fibrous stroma. At the death of the rat, a month later, the
recurrent nodule proved to be highly fibrous, and showed epithelioid
cells in rows and islands in process of compression and atrophy.
We shall now consider a few examples of the fifth generation,
which, as will be recalled, was characterized especially by the appear-
ance for the first time of lymphatic gland metastases. It was in
this generation also that several examples of compression of the
spinal cord occurred. An example of the latter condition will now
be described.
Rat 87 (a) had been inoculated into the muscles of the back, and the growth
had extended toward and invaded the spinal column. Viewed under the micro-
scope the growth was composed in its preserved parts of the large epithelioid
cells forming solid alveoli which had penetrated all the soft parts, substituted
itself for them and surrounded the vertebral column and ribs (Plate VIII, Fig.
16). The spinal nerves at their emergence from the canal were invaded. The
large tumor cells insinuated themselves between the muscle fibers, which were
caused to atrophy, and between the nerve sheaths. The bony parts of the
vertebrae were not destroyed, excepting the spinous processes, in which no bone
remained. The tumor passed to the spinal canal and reached the dura mater,
but the spinal cord was not compressed. It is probable that the paralysis was
produced through involvement of the spinal nerves roots. Rat 198 (d) showed
multiple pulmonary nodules, in which the tubular or acinous arrangement of the
epithelial cells was well marked. The nodules were often as large as a pea and
replaced the pulmonary tissue. The stroma was finely fibrillated and did not
conform to the original architecture of the lung. The pleural nodules were
particularly interesting. While they possessed a denser and more abundant
stroma, they were covered on the free surface with a layer of columnar epitheli-
oid cells proliferating rapidly by mitosis, which on dipping down into clefts and
depressions of the tissue gave rise to typical acinous tubules, as viewed in sec-
tions. A subcutaneous nodule in the same animal showed similar epithelioid
cells proliferating rapidly and forming linear solid strands and not acinous
tubules. There was also in this animal an ingrowth of tumor into the vertebral
column, which invaded the dura mater but did not infiltrate the spinal cord.
The roots of the spinal nerves and the inter-vertebral ganglia were surrounded.
This tumor mass showed interesting variations of type. The older and more
superficial parts were in a state of coagulative necrosis. Immediately next to
them was a mass of solid anastomosing strands of cells pushing into and replac-
ing the muscle fibers, which showed all stages of atrophy. The stroma here
developed was highly fibrous. The large tumor cells surrounding the spinal
nerves developed in the perineurium and produced a heavy collar of cells about
the nerves, so that the main trunks of the nerves were completely degenerated.
The type of tumor cells developed within the fibrous stroma and leading into the
spinal canal was smaller than elsewhere ; the nuclei were darker and the arrange-
Digitized by
Google
30 Studies upon a Transplantable Bat Tumor.
ment was radial about the irregularly anastomosing spaces containing the cell
masses, so that they were made to resemble in a remarkable way the normal cells
lining the narrow spaces. Where the cells could grow more freely in the
muscles and fascia beside the vertebral column, the alveoli became more regu-
larly rounded or branching. Here and there a central excavation had taken
place among the cells, producing a kind of acinous arrangement, to which
the term pseudo-tubular might be applied (Plate IX, Fig. 17). At the outer termi-
nation of the muscles, where the growth might be considered as having been free
to extend, the cells were large and pale, and were rapidly multiplying by
mitosis; they were disposed into columns and made to resemble the form of
growth described in the pleural metastases. Rat 411 (h) showed some inter-
esting features. There was an extensive tumor growth at the root of the lung
involving the great vessels of certain lobes. The vessels were thrombosed and
the corresponding lobes of the lungs were necrotic. The tumor passed at times
into the interior of the vessels, so that epithelioid cells came to lie between the
inner wall and the thrombus in a narrow slit or space, giving rise to tubules
containing large multi-nucleated cells. Elsewhere in the lungs there were
numerous tumor nodules developing about the bronchi and blood vessels, which
penetrated their walls but did not enter the lumena. The tumor cells in these
cases might come to lie just between the endothelium of the vessels and the
epithelium of the bronchi. In these the alveolar formation was solid. The
kidney nodule showed replacement of the parenchyma of the organ by the tumor,
the epithelioid cells of which occupied the pre-formed tubulrs, but the stroma
was newly formed.
We shall consider now the examples of lymphatic gland involve-
ment. As stated, the tendency of the tumor at this time was to
exhibit the structure of carcinoma. The particular type presented
was that of simple carcinoma, in which the alveoli showed great
diversity of form, but contained solid masses of epithelioid cells.
In the course of the fifth generation, however, this simple form
began to show modifications, tending to the acinous arrangement,
and although the two types, the solid and acinous, of alveoli were
fluctuant, yet the former became more and more common in this
and subsequent generations, until it became dominant. We shall
pass to the seventh generation before taking up the consideration
of the nature of the lymphatic gland metastases.
Rat 404 (d) showed, besides the growth in the regional K-mphatic glands,
metastases in the lungs. The subcutaneous nodule had ulcerated and left a
mere capsule of fibrous tissue containing strands of epithelioid cells. It showed
no acini. The lungs showed both disseminated miliary and larger nodules, and
one lung was quite completely converted into tumor. The main blood vessels
leading to this lung were surrounded by tumor and thrombosed, the thrombi
being in process of organization. The bronchi had been either surrounded or
Digitized by
Google
Simon Flexner and J. W. Jobling. 31
invaded by tumor. The pleura also contained nodules. The lungs were partly
bound to the chest wall by the tumor growth, and the areolar tissue under the
heart and lungs was invaded, but the growth, which extended to the lymph
glands in this areolar tissue, had not invaded them. The type of the tumor was
the simple alveolar. On the other hand, the lymphatic glands in the region
of the superficial ulcerated mass had been invaded, some of them almost
entirely replaced by the tumor. The epithelioid cells were pale, some of them
presented hypochromatic nuclei, and they formed solid alveoli (Plate IX,
Fig. i8). The succeeding animals frequently showed a similar replacement of
the regional l3rmphatic glands, either with solid alveoli or a branching type of
the tumor, and in rare instances the tumor passed through the lymphatic glands
into the periglandular adipose tissue.
Rat 729 (d) of the eighth generation showed the manner in which the
lymphatic glands sometimes become infected. The local tumor, of fibrous nature
and containing only a few of the epithelioid cells in small groups, approached
the gland and became fused with the capsule, after which blunt papillary-like
processes, consisting of large epithelioid cells, pushed into the superficial lym-
phatic tissues, which was the starting point of the invasion.
We return for a moment to Rat 401 (d) of the seventh genera-
tion, because of a nodule contained within the auricle of the heart
and penetrating into the cavity of the auricular appendage. The
muscle cells in the region of the tumor were either undergoing
atrophy or had entirely disappeared. The tumor itself was an
alveolar growth of solid type, possessing a rich fibrous stroma.
The ingrowth into the muscle of the heart was accomplished through
buds of multinucleated protoplasm, which, enlarging, produced the
strands and oval alveoli. The cavity of the appendage contained a
fibrinous clot surrounded by hyalin muscle fibres, into which the
tumor had penetrated and produced a tubular formation.
At about this time the tumor showed now and then a much softer
structure than it had previously done, which can best be illustrated
by one or two examples. Rat 808 (d) of the eighth generation
possessed a tumor the size and shape of an almond, that was com-
posed of cell masses made up of large pale epithelioid cells arranged
either in solid alveoli or in tubules, showing lumina. The cortical
portion was made up of living cells, while the central part, except
for the tissue immediately about the blood vessels, was the seat of
a soft necrosis, quite different from the hyalin necrosis usually
observed. In this necrosis, which resembled an anaemic infarction,
the cell forms were still present, but distorted and discolored and
Digitized by
Google
32 Studies upon a Transplantable Bat Tumor.
mingled with fragmented nuclei. This particular form of the tumor
became progressively more common, and led to the typical adenoma
that constituted the final form in which the tumor occurred. In
different examples there was some difference in arrangement of
epithehoid cells and in relation of cells and stroma, but the proper-
ties were generally preserved. What was particularly impressive in
this tjTpe of the tumor were the frequent large mitoses in the epithe-
lioid cells.
Rat 743 (d) was another example of this type of tumor, which
was even more perfectly adenomatous. The mass was several centi-
meters long, oval in shape, and possessed a fibrous capsule. The
cortical layer, still living, was about two millimeters in thickness,
while the central part was necrotic and soft, but still showed small
islands of living tissue about the capillary and larger blood vessels.
There was little or no stroma, and the bulk of the growth consisted
of epithelioid cells, proliferating very rapidly and, therefore, show-
ing numerous mitoses. The individual cells were ill-defined, but
multinucleated bands and rows of cells had been produced. At the
outermost edge of the tumor the epithelioid cells were narrow and
often linear, but further inward they were wider, formed oblong
masses and frequently showed a peripheral circle about a lumen.
There was much curving and bending of the epithelioid cells, so
that acini of a variety of shapes were produced. In the islands of
tissue left preserved about the vessels in the interior, a tubular,
acinous arrangement nearly always was present (Plate X, Fig. 19).
To illustrate the fluctuations which were still going on at this
period. Rat 1280 (i) of the ninth generation will be used as an
example. Besides a nodule in a lymphatic gland, which perforated
the capsule and grew into the surrounding areolar tissue, the orig-
inal nodule was composed of simple alveoli containing pale cells and
showing numerous mitoses. The amount of degeneration was not
excessive, and no tubules or acini were produced. Two other points
may be mentioned: one of the large superficial nerves was sur-
rounded (Plate X, Fig. 20), and a number of mammary gland
ducts were included in the growing tumor. This latter condition
was by no means uncommon when the implantations were made in
the region of the mammary glands of female rats. Frequently
Digitized by
Google
Simon Flexner and J. W. Jobling. 33
these ducts persisted in the tumor and showed little change. At
other times their proper membranes were destroyed and the ducts
themselves disintegrated. Rat 2402 (0) of the ninth generation
showed a lymphatic metastasis consisting of simple alveoli and con-
taining cysts. From the wall of one such cyst an outgrowth of
papillomatous character had taken place (Plate XI, Fig. 21). Rat
2668 (q) showed a combination of anastomosing branching cell-
strands and small solid alveoli filled with pale cells and showing
many mitoses. There were lung nodules composed of small solid
alveoli of more irregular size than the air vesicles, indicating that
they had been newly formed. A new formation of stroma had
taken place, and irregular giant cells of the megacaryocytic type
were also present (Fig. 7). A considerable rarity was the occur-
rence of a metastasis in the wall of a vein ; and another, a growth in
the lumen of a lymphatic vessel (Plate XI, Fig. 22).
Rat 2805 (/) of the tenth generation showed a local lesion of the
ordinary alveolar type and a growth into the kidney extending from
the cortex to the pelvis and invading the renal vein. The tubules
had as usual been invaded and obliterated before the glomeruli.
The tumor passed to the pelvis and came to lie beneath the pelvic
epithelium, which was slowly replaced, but before the replacement
was completed the epithelial membrane of the pelvis had undergone
a considerable hypertrophy, through which it was increased several
times in thickness. Ultimately it became necrotic. The growth of
the tumor into the connective tissue of the pelvis caused it to sur-
round the arteries and veins. The former escaped invasion and
occlusion. One of the latter was invaded and became thrombosed.
Rat 1648 (a) of the eleventh generation was interesting as having
shown for the first time a lymphatic gland metastasis of the typical
adenomatous type, although there were still combined with it solid
alveoli, while Rat 2253 (c) showed the superficial tumor to be of
the soft, pure adenomatous or acinous type. (Plate XI, Fig. 23,
Plate XII, Fig. 24, Plate XVI, Photograph 6.)
The succeeding generations were made up of tumors presenting
chiefly the adenomatous form, but now and again there still appeared
tumors of firmer consistence, containing solid alveoli and fibrous
stroma, so that up to the present time the typical adenomatous type
Digitized by
Google
34 Studies upon a Transplantable Bat Tumor.
has not become firmly established. Thus Rat 3762 of the twenty-
first generation (d) showed a nodule the size of a pea in the sub-
cutaneous tissue, composed of small acini which were located chiefly
centrally, and small, solid, elongated and oval alveoli making up the
larger part and contained within a fibrous stroma. All the cells
were pale and undergoing rapid mitosis. The stroma was remark-
able for its vascularity. At the extending edge the reticulated tissue
contained many small round and spindle cells. Rat 3855 of the
twenty-third generation (c) presented a subcutaneous nodule, the
central part of which was of typical adenomatous, and the cortical
part of fibrous and small alveolar structure. This tumor also con-
tained innumerable mitoses. The adjacent striped muscle was being
invaded by solid cylinders of the pale cells, which were breaking
up into fragments the muscle fibers included within the alveoli or
acini. There was no remarkable increase of muscle nuclei, but the
morsels of muscular tissue were surrounded by epithelioid cells and
were undergoing digestion, although they were not included, prop-
erly speaking, within cells. In a similar manner Rat 3983 of the
twenty-sixth generation (a) showed a local tumor made up of acini
and solid alveoli, which was invading the voluntary muscle. The
cells composing the tumor at the point of junction with the muscle
exhibited a diffuse alveolated growth in which the distinction between
epithelioid cells and stroma was indistinctly marked. On the other
hand, the muscle was undergoing the same form of fragmentation
and solution as has been just described. Rat 3901 of the twenty-
fourth generation (c) contained a subcutaneous nodule, which,
while considerably degenerated, was an excellent example of a
branching or papillary adenomatous tumor. The stroma consisted
of a delicate fibrous tissue arranged in folds and projections, show-
ing secondary branching, which were everywhere covered with a
layer of cells, one or more in thickness, of the epithelioid type, and
sometimes of columnar form. At the periphery of the nodule next to
the capsule, the branching was less obvious, while acini of the typical
form were present. Specimens of this tumor stained by Mallory's
phosphotungstic acid method showed extremely well the differentia-
tion between stroma and large cells, and brought out strikingly the
branching character of the large papilliferous alveoli. Rat 3917 of
Digitized by
Google
Simon Flexner and J. W. Johling. 35
the twenty-fourth generation (d) presented another example of the
branching papillomatous tumor. In this instance the stroma formed
merely a small part of the mass of the tumor, so that by far the
greater part was composed of epithelioid cells. It was towards the
center of the mass, where the degeneration was advanced, that the
papillomatous character was especially pronounced, while in the
periphery the branching was more confined, and large solid alveoli,
as well as smaller acini, were produced. Towards the center the
islands of epithelium were preserved, in the midst of a soft degen-
eration, but these preserved cells in all cases surrounded the small
vessels, the lumina of which were open. Everywhere in the tumor,
and even in the islands mentioned, mitosis was going on most actively.
Rat 3925, twenty- fourth generation {e), was an example of a sub-
cutaneous tumor much degenerated, showing both simple acini and
simple solid alveoli, and transitions of one to the other. Rat 3970,
twenty-fifth generation, showed a subcutaneous nodule composed
partly of acini and partly of solid alveoli, both containing pale cells.
The lung of this rat contained a large nodule. This was necrotic
with the exception of the extending edge, which passed without
sharp demarcation into the adjacent congested lung tissue. The
extension was made by means of blunt, solid outgrowths of the main
tumor mass into the adjacent air vesicles, and there was an abun-
dant increase in the interstitial tissue of the lung. The diffuse char-
acter of the proliferation made it somewhat difficult to determine
the tumor nature of the nodule. Rat 4080 of the twenty-eighth
generation (c) showed a circumscribed pulmonary metastasis com-
posed of pale cells occupying the pulmonary alveoli and completely
filling them. There were local areas of degeneration just begin-
ning, with emigration of leucocytes and fragmentation of nuclei,
but there was absence of typical acinous arrangement.
INTRA-TESTICULAR INOCULATION AND RETROGRESSION.
In the course of the inoculations several attempts were made to
produce intra-testicular growths, with the idea of causing a general
lymphatic dissemination of the peritoneum through that means. In
all the experiments except one the result was a circumscribed necrosis
Digitized by
Google
36 Studies upon a Transplantable Rat Tumor.
of the testicle, with more or less new formation of fibrous tissue,
but no tumor growth. In one experiment tumor nodules were pro-
duced. This was in the eighth generatioa The plan was to make
emulsions of the tumor in salt solution and to inject them into the
testicle. In the successful experiment two rats developed tumor
nodules the size of small shot, just beneath the tunica of the testis
and penetrating between the tubules. It was only in the periphery
of the nodule that preserved, multiplying tumor cells, lying in a
fibrous stroma, were observed. The centers even of these small
nodules were necrotic, and consisted partly of degenerated tubular
elements. There was no dissemination of the tumors.
Attention has repeatedly been drawn to examples of tumors which
showed an unusual amount of fibrous tissue. Fibrous tissue is pro-
duced in two ways : ( i ) It is constantly produced at the edge of the
growing tumors, in the course of the formation of the stroma and
the capsule. It, however, is also produced in the interior, as a
result of degeneration of the epithelioid elements. In growing
tumors the central fibrosis comes to be quickly associated with coag-
ulative necrosis of the tumor, which affects first the remains of the
epithelioid cells and next the fibrous tissue. Much of this fibrous
tissue is hyalin in structure, so that its true character may be more
or less masked. (2) The second set of conditions producing fibrous
transformation of the tumor is the one which we wish especially
to refer to in this place. It occurs in the course of the retrogression
of the tumors, and consequently is found in small as well as in larger
nodules. In these cases the fibrosis tends to be unassociated with
the coagulative necrosis. The fibrillation of the tissue also appears
to be preserved and the epithelioid elements to be contained in con-
tracted, narrow, often linear spaces within the tissue. The number
of these cells is greatly reduced, being smallest in the central parts
where the fibrosis is oldest. By compression they are eventually
entirely destroyed through atrophy, in the course of which the nuclei
become contracted and pyknotic, and the protoplasm, which takes
on a deeper eosin-staining than it does normally, is reduced to a
mere rim. It is the peripheral portion which contains the least
fibrous tissue and the most epithelioid cells, but there is a striking
absence of mitosis. Briefly expressed, retrogression is associated
Digitized by
Google
Simon Flexner and J. W. Johling, 37
with an increased fibrosis, a reduced mitosis, and a gradual and ulti-
mately complete shrinkage of the tumor. The removal is not,
apparently, through the action of leucocytes. At least, there does
not go on any gross destruction by means of these cells. In the end
the fibrous tissue is as completely removed as the epithelioid ele-
ments, and nothing remains in the local tissues to show the presence,
at a previous time, of the tumor.
THE ELEMENTS FROM WHICH THE TUMORS ARE DERIVED.
In order to determine roughly the elements upon which successful
implantation of the tumor depended, several methods were em-
ployed, the results of which can be summarized briefly.
Emulsions of the tumor prepared in salt solution were filtered
through unglazed porcelain, and the filtrate injected into the abdom-
inal cavity and beneath the skin of the rat. In no case did a tumor
develop.
Emulsions of the tumor in salt solution were filtered through
sterile filter paper and injected into rats in the same manner, but in
no case did a tumor develop.
Emulsions of the tumor in salt solution were passed merely
through sterile gauze, and the fluid, carrying minute fragments of
the tumor, was injected in the same manner into other rats, but
never gave rise to tumors. On the other hand, it has been proven
that similar small fragments of the tumor, when not injured by the
medium in which they are suspended, can give rise to the production
of tumors. This has been shown by the successful results of the
inoculations of ascitic fluid containing microscopic masses of tumor
cells (see page 21).
With a view of excluding, or, indeed, of confirming the presence
of microscopic organisms in the tumor, sections of young growing
tumors were stained with all the usual aniline dyes in the various
ways employed to show bacteria or protozoa, were impregnated with
silver nitrate according to Levaditi's method of staining spirochaetae,
and were examined in the fresh condition with the ordinary high
powers of the microscope and with the special apparatus for dark
field illumination, but nothing was ever seen that was suggestive of
any other structure than the proper cellular elements of the tumor.
Digitized by
Google
38 Studies upon a Transplantable Bat Tumor.
Aerobic and anaerobic cultures were made from young undegen-
erated and older degenerated tumors, but they never yielded any
growth of bacteria.
THE ACTION OF PHOTO-DYNAMIC SUBSTANCES ON THE TUMOR.
Experiments showed that when bits of the tumor of some size
were kept immersed for a brief period in physiological salt solution,
tumors could still be reared from them, but they also showed that
the minute fragments which would pass through the meshes of ordi-
nary gauze were no longer able to yield tumors when exposed to the
action of this solution. The failure in the latter experiments was
attributed to the poisonous effects of the sodium chloride; and it
was proved that Ringer's solution, in which these poisonous effects
are balanced by the use of the calcium, was less injurious to the
tumor than salt solution alone. A better, because less injurious,
fluid for suspending the tumor proved to be rat serimi heated to 54°
C, but in making the transplantations the fragments of tumor which
were introduced were not permitted to come into contact with any
extraneous fluid whatever. In this way the best results were
obtained.
An effort was made to influence the growth, persistence and
transplantability of the tumor through the action of photodynamic
agents. We believed at one time that we had secured such an action
by means of eosin and Bengal red, but later study of the action of
these elements led us to the conclusion that the early results which
we had obtained and supposed to be due to the action of the photo-
dynamic anilines was attributable to other causes. We shall give
a table illustrating the irregularity of the results of transplantation
of a single specimen of the tumor, illustrating particularly the great
variation in the number of retrogressions taking place in a series.
The method was to expose minute fragments of the tumor to the
action of the various solutions given in the table. It was intended
that the exposure should be one hour, but the large number of
animals to be inoculated required that the exposure extend from one
to three hours, since a part of the inoculations were not completed
until the expiration of that time. However, the light exposures
were all limited to one hour.
Digitized by
Google
Simon Flexner and J. W. Jailing.
39
1 No.
Tumors, Per Cent.
Retrogressions. Per Cent.
Controls
lo
10 or ICO
3 or
30
NaCl : light
lo
10 or loo
3 or
30
NaCl : dark
ID
10 or lOO
4 or
40
NaCl eosin i-iooo : light
lO
9 or 90
3 or
U
NaCl eosin i-iooo : dark
lO
9 or 90
8 or
NaCl Bengal i-iooo : light
lO
10 or 100
10 or
100
NaCl Bengal i-iooo : dark
9
8 or 88
I or
12.5
Ringer's : light
lO
7 or 70
6 or
85.7
Ringer's : dark
lO
9 or 90
2 or
22
Ringer's eosin i-iooo : light
lO
10 or 100
3 or
30
Ringer's eosin i-iooo : dfurk
9
9 or 100
3 or
33
Ringer's Bengal I- 1 coo : light
9
9 or 100
2 or
22
Ringer's Bengal i-iocx) : dark
lO
9 or 90
2 or
22
Serum : light
ID
10 or 100
2 or
20
Serum : dark
7
7 or 100
2 or
28
Serum eosin i-iooo : light
lO
10 or 100
5 or
50
Serum eosin i-iooo : dark
lO
10 or 100
4 or
40
Serum Bengal i-iooo : light
10
or 90
8 or 80
I or
II
Serum Bengal l-iooo : dark
ID
5 or
62
An explanation of the table brings out the fact that the tumor
under the influence of the anilines tends to give a larger number of
retrogressions than in the absence of the anilines, and also that expo-
sure to sunlight for one hour is possibly to some small extent inju-
rious, but no general deductions can be made from the experiment,
and we publish this final table in order to set ourselves right on this
point.
MODE OF DEVELOPMENT OF THE TRANSPLANTED TUMOR.
Leo Loeb, and later Bash ford, and since then others, have made
careful observations on the manner of development of the trans-
planted fragments of tumors. We have ourselves studied the man-
ner in which the fragments of this tumor undergo development in a
new animal. Indeed, we have studied the changes in the trans-
planted fragments which occur not only in the rat but also in the
mouse and the guinea-pig. We owe to Ehrlich the important obser-
vation that a tumor fragment introduced into a species from which
it was not developed, as from a mouse into a rat, would undergo a
limited development in the new species, but would never attain a
considerable size or become organically attached to the individual
of the new species, or be transplantable from one to another of this
new species. On the other hand, such living fragments can be
Digitized by
Google
40 Studies upon a Transpla7itable Rat Tumor.
transferred successfully at some periods from the new to the original
species, and then be transplanted further.
We have found in common with Loeb and Bash ford that the
greater part of the tissue transplanted undergoes necrosis, and that
the new tumor is produced from a small number of surviving cells.
Our observations were made with the usual fragments transplanted
by us and also with a somewhat finer emulsion (made by commi-
nuting the living cortical parts of the tumor by means of a Haaland
grinder) deposited beneath the skin in the region of the axilla. A
series of rats, mice and guinea-pigs were inoculated on the same day
with fragments and emulsion from one source. They were chloro-
formed at intervals of twenty- four hours ; the implanted tumor wnth
the surrounding tissue was removed, preserved and hardened in
Zenker's fluid, and sections stained in hematoxylin, eosin and iron-
hematoxylin were prepared.
FRAGMENTS IN THE RAT.
The fragments of tissue in the emulsions were several times
smaller than the morsels which were introduced by means of the
inoculating needle ; fewer of them survived, and their development
was on the whole much less rapid. The essential process of growth
was, however, identical in the two series.
At the end of twenty-four hours considerable portions of the
transplanted tissue fragments had become necrotic. The tissue of
the fragments, excepting the surviving cells, to be described, pre-
sented appearances similar to those which tissue permitted to undergo
autolysis at the body temperature outside the body would be expected
to show. The interstitial tissue showed few deformed and frag-
mented nuclei; the fibers were indistinct and more or less fused
together ; and the epithelioid cells occupying small spaces in the latter
tissue were shriveled, with irregular outline, the protoplasm staining
deeply in eosin, the nuclei of irregular form and unequal staining
powers, and in process of disintegration. The fragments were
everywhere, as was to be expected, surrounded by a fibrinous and
leucocjrtic exudate, and a certain number of leucocytes had pene-
trated within them. Within certain of the fragments a number of
Digitized by
Google
Simon Flexner and J. W. Jobling, 41
epithelioid cells had survived. These cells might be at the periphery
or they might be in the interior of the fragments, sometimes single
and sometimes several in a given space. They differed greatly from
the degenerating and disintegrating cells, particularly in respect to
the behavior of their nuclei, which retained the normal form and
distribution of chromatin, although in point of intensity of staining
they tended to be hypochromatic. For the most part the cells
appeared merely to be survivals, but now and again nuclear changes
suggestive of mitosis were met with. However, in view of the fact
that the chromatic filaments were somewhat irregular in thickness
and arrangement, it cannot be said that those changes actually led
to cell multiplication at this early period. It should be mentioned
that in some fragments the old blood vessels still remained morpho-
logically unchanged at the end of this period.
At the expiration of forty-eight hours the effects of the trans-
plantation were still more obvious. The stroma and the greater
part of the epithelioid cells and the blood vessels in the transplanted
morsels had all progressed still further in respect to the disinte-
grative changes. On the other hand, the surviving epithelioid cells
had, either by contrast or actuality, become more sharply demarcated
from the surrounding elements. Where they were contained within
the interior of the fragments, their number had not greatly increased,
although apparently there had already been some increase in num-
ber; the nuclei had become richer in chromatin, the surrounding
protoplasm clearly outlined, and mitosis had not only become more
frequent than in the twenty-four hour period, but the chromatin
filaments presented a regular arrangement and uniform thickness
suggestive of normal mitosis. At the periphery of some of the
fragments the epithelioid cells had clearly multiplied, so that they
had come to form a superficial layer, one, two or three cells deep in
some places. These cells lay in contact with the inflammatory exu-
date, and some of them apparently had wandered into the exudate.
Mitoses also occurred in them, and even in separated cells surrounded
by the exudate. Thus far no organic connections had taken place
between the fragments and the surrounding tissue, although in the
latter an active proliferation and new formation of capillaries had
begun. On the other hand, elongated cells of fibroblastic or angio-
Digitized by
Google
42 Studies upon a Transplantable Rat Tumor.
blastic nature had begun to enter the fragments at the periphery,
adjoining the host tissues.
After three days the conditions had altered greatly. The frag-
ments had not increased perceptibly in size, but those lying adjacent
to the host tissues had become attached through capillary connec-
tions with the tissues of the living animal. The capillaries had
pushed in various directions and some of considerable size had
already formed channels in the interior. The medium in which
they lay was apparently nothing else than the degenerated original
stroma, and there had been no unmistakable restoration of living
interstitial cells. All the cells which had wandered in were in con-
nection apparently with the capillaries. On the other hand, the
epithelioid cells had increased greatly, mitosis having gone on rap-
idly. The cells now formed small groups chiefly, although some-
times single ones were separated from the others. The active changes
in the vascularized fragments were greatly in advance of those not
already vascularized. Moreover, those fragments or portions of
fragments in which the epithelioid cells had not begun to multiply,
nor the capillaries to penetrate, had by this time been invaded with
large numbers of leucocytes, which were softening and dissolving
the necrotic tissue. Finally, in certain fragments there had ap-
parently been beginning multiplication of the epithelioid cells, which
had not become vascularized, and which were undergoing a secon-
dary degeneration and death.
At the end of the fourth day the conditions in the vascularized
peripheral fragments represented merely progression of the tliird.
In other words, the vascularity having increased, these fragments
were now entirely occupied by capillary vessels and proliferated
epithelioid cells, while the fragments which were more interior were
less advanced in this condition, and might show mitosis of epithe-
lioid cells still going on, although no capillaries were yet visible in
the stroma, which was quite completely necrotic. At the same time,
occasional islands of epithelioid cells remained alive and were pro-
liferating slowly.
The subsequent progress was rapid. The groups of epithelioid
cells continued to increase in number and to expand in size, and
coincidentally a fibrillated tissue appeared between them and the
Digitized by
Google
Simon Flexner and J. W. Jobling. 43
capillaries. At the end of seven days the fragments had become
completely vascularized, all of the evidences of acute inflammation
had disappeared, the volume of the fragments had increased several
fold, and they had formed an indissoluble union with the tissues of
the host. At the junction of the two, a small-cell granulation tissue
formed a sort of capsule. The positions of the original fragment
could still be made out through the remains of a central, paler, par-
tially degenerated, and not completely substituted mass of epithelioid
cells and young capillaries, with here and there polynuclear leuco-
cytes and beginning fibrillated tissue. What was particularly strik-
ing was the close union of the epithelioid cells and the young con-
nective tissue. This union was so close that alveolation in the
proper sense of the term could not be spoken of. In addition, there
occurred not infrequently multinucleated protoplasmic masses giving
rise to giant cells and large single cells with giant nuclei.
By the ninth day the differentiation between the fibrous stroma
and the epithelioid elements had progressed so far that as the con-
nective tissue became more fibrous it left the epithelial cells more
sharply demarcated. On the eleventh day this demarcation was
still more pronounced, although the condition was not as yet one
of definite alveolation. At this time the remains of the original
fragment occupying the central portion of the enlarged nodule was
still evident. It consisted of a hyalin mass enclosing small groups
or single epithelioid cells not in the best state of preservation. On
the fourteenth day the nodule was still further increased in size, the
connective tissue had become more pronouncedly fibrous, but the
epithelioid cells were not so strictly confined as to produce a typical
alveolated structure. On the other hand, on the fifteenth day the
fragment was quite characteristically alveolated.*^ The connective
tissue now formed a definite fibrous stroma enclosing alveoli, chiefly
solid but sometimes presenting spaces and, therefore, acinous, of
epithelioid cells, which were multiplying rapidly by mitosis. These
alveoli were shown, especially in the preparations stained with iron-
hematoxalin, to be of irregular size and form and everywhere inter-
sected by the fibrous stroma. In that respect at this period the
•This change is one of degree only and would doubtless vary in different
experiments.
Digitized by
Google
44 Studies upon a Transplantable Rat Tumor.
alveolation was not by any means so regular as was found in some
later nodules and as is the rule in simple carcinoma in man.
FRAGMENTS IN THE MOUSE.
The results of the introduction of fragments of the tumor under
the skin of the mouse were in many respects, for the early period,
similar to those in the rat. The first effect was, of course, a cer-
tain amount of inflammatory exudation, but this amount was not
greater than that which took place in the rat. The changes in the
fragments were also similar, which is to say that the greater part
of the tissue composing them degenerated. But what remained
alive at the end of twenty- four hours was a certain number of
epithelioid cells on the surface or in the interior of the fragment.
At the end of the first twenty-four hour period mitosis had not
been noticed in the host cells, but the stroma and many of the
epithelioid cells were in a state of complete necrosis, and the
resistant epithelioid cells had begun to show mitosis. The forma-
tion of angioblasts and the organic connection with the mouse of
the fragments also occurred on the third day, at which time there
had been a great proliferation by mitosis of the epithelioid cells,
particularly those lying immediately next to the tissues of the mouse,
which were united to the fragments, although the mitosis was not
limited to those cells. The increase in the epithelioid cells on the
fifth day was considerable, and they formed groups within the
homogeneous residue of the stroma, which- was now in process of
invasion by capillaries. On the seventh day the new vessels had
penetrated deeply into the fragment ; there was a striking hyperemia,
and the epithelioid cells were still in process of multiplication. On
the eleventh day the healing process was still more advanced, but
remains of the epithelioid cells, which were, however, inactive, were
still present. On the fifteenth day the vascularization was still
more advanced, and had reached almost to the center of the frag-
ments, which were occupied by masses of leucocytes undergoing
fragmentation, but the specimens showed that multiplication by
mitosis was still going on actively in the epithelioid cells, especially
in the vicinity of the tissues of the host. The absence, therefore,
Digitized by
Google
Simon Flexner arid J. W, Jobling. 45
of active mitosis in the two previous periods was the result not of
the exhaustion of the tissue but of accidental circumstances in con-
nection with the transplanted fragments. Although the latest
period in which active mitosis was discovered in the fragments
transplanted to the mouse was fifteen days, the large number of
them contained in one series of specimens of that period indicates
than they might still be found under favorable circumstances even at
a later period.
We have transferred fragments which have remained in the
mouse for different periods back to the rat and have succeeded in
rearing these fragments as late as eight days after their implanta-
tion in the mouse.
FRAGMENTS IN THE GUINEA-PIG.
Fragments transplanted to the guinea-pig aroused a somewhat
greater inflammatory reaction and exudation than in the other
animals mentioned. The changes which took place in the trans-
planted fragments did not differ essentially from those already
described. While the great mass of the tissue died and showed the
changes of autolysis, groups of epithelioid cells in the interior or
near the surface survived. At the end of the first twenty- four hour
period these surviving cells were still quiescent, while at the end of
forty-eight hours they had changed significantly : they had become
sharply demarcated from the rest of the tissue, were more numerous
than they had been in the previous twenty-four hours, and were
showing active mitotic division. Indeed, this activity was more
marked than it had been under similar circumstances, either in the
rat or the mouse. The multiplication was taking place in the
interior of the fragments, as well as on the surface. At the latter
situation the new cells had already formed a layer several deep.
There was still activity, but it was less in these cells at the end
of the third day, at which time there were fewer organic connections
with the host than had been noted in the other animals, and many
fewer new blood vessels had formed. On the seventh day the
vascularization was well advanced, but not equally about all the
fragments. About some the exudation was very great and had
Digitized by
Google
46 Studies upon a Transplantable Bat Tumor.
penetrated into the interior. Karyokinesis was, however, still going
on in the epithelioid cells, although not actively, and particularly
at the periphery of the fragments. The epithelioid cells in the
interior were reduced in numbers, and were degenerating. On the
ninth day the fragments had been greatly reduced in size and had
become much more disintegrated, so that the demarcation of the
epithelioid cells was uncertain.
SUMMARY.
The rat tumor described is to be regarded probably as an em-
bryoma located in the seminal vesicle, from whose histological ele-
ments it was partially derived.
Originally the tumor was composed apparently of a simple tissue
of imperfectly organized structure, but containing certain glandular
elements derived in part from the seminal vesicle and in part from
other included epithelial organs.
The included glandular elements had already undergone in the
original tumor a carcinomatous development, although the carcin-
omatous moiety was not disseminated throughout the mass, but was
confined to portions near its peripheral boundary, adjacent to the
wall of the vesicle.
In the course of transplantation the simple and imperfectly
organized structure of the tumor became substituted by a more
complex and perfectly organized tissue, which bore resemblance to
simple alveolar carcinoma.
The tumor at this time not only developed in the subcutaneous
and intra-muscular tissue and the abdominal cavity, according as
the grafts were made, but it gave rise to metastases through the
blood current in the lungs and kidneys.
In the fifth generation the simple carcinomatous structure began
to be replaced by a tissue resembling adeno-carcinoma, and coin-
cidentally metastases appeared in the superficial and rarely in the
deeper, distant lymphatic glands.
The adeno-carcinomatous form of structure became progressively
more established in the succeeding generations, in process of which
the stroma of the tumor changed to a delicate fibrillated tissue sup-
porting the acini, and diflfering materially from the rich fibrous
stroma that usually supported the simple alveoli.
Digitized by
Google
Simon Flexner and J. W. Jobling. 47
The tumor had not become established as an adeno-carcinoma of
the pure type by the twenty-eighth generation, at which time simple
alveoli still appeared, but the stroma of the tumors bearing
the latter structure was relatively delicate, and the consistence of the
tumors relatively soft, in contradistinction to the earlier type, which
was much firmer.
The degenerating tumors of the earlier type showed much fibrous
tissue of a hyalin quality, while those of the later type showed
coagulative necrosis of a soft quality.
The metastases partook more of the nature of simple carcinoma
than of adeno-carcinoma, even in the later type. The tumor cells
composing the metastases tended to be substituted for the epithelial
cells of the lungs and kidneys and the lymphoid cells of the lym-
phatic glands, and merely to occupy the already prepared stroma
of these organs. Hence the structure of the organs in the metas-
tases was not greatly disturbed. When a metastasis appeared
in the pleural membrane, in which fibrous tissue was readily pro-
duced, the acinous type of tumor tended to appear, and in the course
of the invasion of the cardiac muscle and other solid tissues, acini
also tended to be produced.
The tumor presented a high degree of invasive power, and its
growth was not interrupted by muscle, fascia, bone or cartilage.
The dilated esophageal extremity of the stomach tended to become
invaded. When the tumor reached the mucous membrane, ulcers
developed, and implantation to the opposite surface of the mucosa
sometimes occurred.
The retrogression of the growing tumors was accomplished
through atrophy and degeneration of the epithelioid cells contained
in the tumor and an increasing degree of filn-osis of the stroma.
Ultimately all parts of the tumor disappeared completely. Nothing
was discovered to indicate that the metastases ever underwent
changes of the nature of those leading to retrogression of the
superficial nodules.
Although the metastases were such as were produced chiefly
through the blood current, it was most rare to find masses of tumor
Digitized by
Google
48 Studies upon a Transplantable Rat Tumor.
cells in the blood vessels, unless they had penetrated them second-
arily from without.®
The manner of development of the fragments of the transplanted
tumor showed the large epithelioid cells to be the proper elements of
the tumor. All other tissue elements succumbed, and the reproduc-
tion of the tumor depended upon the proliferation by mitosis of
these cells, the stroma for which was supplied by the tissues of the
new host.
There was no essential difference in the manner of survival of
grafts introduced into rats, mice and guinea-pigs, except that in
the last two species the process of multiplication of the epithelioid
cells came to an end in a few days, in spite of the vascularization
of the transplanted fragments, which at first promoted proliferation.
The number and location of fragments surviving in the rats
depended in part on the rapidity and degree with which the vas-
cularization took place. The first stages of proliferation did not
depend upon the vascularization, but unless new vessels had entered
the fragments by the third day, the epithelioid cells already produced
tended to degenerate rather than to proliferate further.
A highly glandular organ such as the testicle proved not to be
very suitable for the tumor implantations.
Salt solution proved to be injurious to the tumor tissue, especially
when it was in a state of fine division. The tumor was acted on
injuriously by sunlight, and this injurious action appeared to be
somewhat intensified by the presence of certain photo-dynamic
anilines.
Filtered tumor emulsions and even strained tumor emulsions in
salt solution did not produce new tumors upon inoculation, while
small aggregates of the tumor cells, such as occur in ascitic fluid,
did set up the formation of new tumors.
No microorganisms were recognized in the tumor, and no bac-
terial cultures were obtained from non-ulcerated tumors.
• We arc indebted to Miss Menten for a very careful study of large numbers
of serial sections of the lungs of rats, in which metastases were present, and in
which no visible metastases could be seen. She succeeded in a single instance
in finding a small mass of tumor cells within a blood vessel, unassociated with
a developed metastasis, and this was in a lung already the seat of metastatic
nodules.
Digitized by
Google
MONOGRAPH NO. 1. PLATE I.
Fig. I. Fir. 2.
Fig. 3.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE II.
Fig.
Fig. 5. Digitized by
Google
Digitized by
Google
IVONOGRAPH NO. 1. PLATE III.
7/0 mm. 1
Fig. 6.
Fir.. 7.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE IV.
Fig. 8.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE V.
Fig. 9.
Fig. 10.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE VI.
Fig. II.
Fig. 12.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1.
PLATE VII.
»^
V,
'l^nini
lf¥L
^J^
^.*?5LJ^J^@ai
Fig. 13.
Fig. 14.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE VIII.
Fig. 15.
t'lr.. 16.
Digitized by
Google
V
Digitized by
Google
MONOGRAPH NO. 1. PLATE IX.
Fig. 17.
Fir,. 18.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE X
Fig. 19.
Fig. 20.
Digitized by
Google
Digitized by
Google
lONOGRAPH NO. 1. PLATE XI.
Fig. 21. Fig. 22.
Fig. 23.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XII.
Fig. 24.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XIII.
PHCTOGRAI'H I.
Digitized by
Google
Digitized by
Google
PONOGRAPH NO. 1. PLATE XIV.
Photograih 2 Photograph 3.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XIV.
PHOTOGRAI'H 2 PHOrCHIRAPH 3.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1.
PLATE XV
PHOTOIRAPH 4. PliOTOGRAPH 5.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XVI.
Photograph 6.
Digitized by
Google —
Digitized by
Google
Simon FUxner and J. W. Jobling. 49
EXPLANATION OF PLATES.
Plate L
Fig. I. No. 3 eye-piece; 6 objective. Showing the relation between the
spindle and polymorphous cells occurring over considerable areas of the tumor.
Fig. 2. No. 3 eye-piece; 6 objective. Glandular tubule included in the tumor
undergoing a typical proliferation or being dissected by ingrowth of the tumor.
Fig. 3. No. 3 eye-piece; 3 objective. Dilated glandular tubule enclosed in
hyalin connective tissue adjacent to alveoli enclosing epithelial cells, probably
derived from the tubule.
Plate IL
Fig. 4. No. 3 eye-piece; 3 objective. Junction of the seminal vesicle, show-
ing dilated glandular spaces, with the tumor, and typical carcinomatous alveoli
possessing lumena.
Fig. 5. No. 3 eye-piece. 3 objective. Branching tubular formation or duct
adjacent to a glandular mass possessing structures of a serous or mucous gland.
Plate IIL
Fig. 6. No. 3 eye-piece; 6 objective. Irregular carcinomatous alveoli em-
bedded in the stroma, containing spindle and other cells resembling the epithe-
lium-like polymorphous cells of the alveoli.
Fig. 7. No. 3 eye-piece; 6 objective. Pulmonary tumor nodule of alveolar
structure, showing a mononucleated giant cell of the megacaryocytic tj-pe.
Plate IV.
Fig. 8. No. 3 eye-piece; 6 objective. Invasion of the heart wall and coronory
sinus by the tumor.
Plate V.
Fig. 9. No. 3 eye-piece; 6 objective. Stomach opened to show the ulcerated
tumor masses in the esophogeal segment. The elevated, undulating white line
shows the demarcation between the esophogeal and gastric mucosa.
Fig. id. No. 3 eye-piece; 3 objective. Subcutaneous tumor nodule showing
the relation of the young alveoli to the stroma. A definite demarcation separates
the tumor from the cutis.
Plate VI.
Fig. II. No. 3 eye-piece; 6 objective. Showing the usual structure of the
tumor in the stage of carcinoma simplex.
Fig. 12. No. 3 eye-piece; 6 objective. From the edge of a recent rapidly
growing graft showing the original development of spindle cells and small
included masses of polymorphous cells which have not yet arranged themselves
into d'efmite alveoli.
Digitized by
Google
studies upon a Transplantable Sat Tumor.
Plate VII.
Fig. 13. No. 3 eye-piece; 6 objective. Hyalin transformation of the tumor
in the course of . retrogression showing the persistence of small groups and
masses of the polymorphous cells within the hyalin connective tissue.
Fig. 14. No. 3 eye-piece; 3 objective. Invasion of the splenic vein by the
tumor which presents a typical tubular alveolar structure. The vein is occluded
by a thrombus which is itself becoming invaded with tumor acini.
Plate VIII.
Fig. 15. No. 3 eye-piece; 3 objective. The invasion of the submucous and
subepithelial layer of the esophageal segment of the stomach by the tumor which
shows distinct alveolation. Beginning degeneration of the superficial epithelial
cells.
Fig. 16. No. 3 eye-piece; 3 objective. Invasion of the vertebral column.
The alveolar tumor in the section is adjacent to the cartilages and bony structures.
Plate IX.
Fig. 17. No. 3 eye-piece; 6 objective. Pseudo-acinar excavation of the
alveoli.
Fig. 18. No. 3 eye-piece; 6 objective. Invasion of the regional lymphatic
glands by the alveolated tumor. The stroma which is slight is furnished
by the gland.
Plate X.
Fig. 19. No. 3 eye-piece; 3 objective. Atypical carcinoma simplex in which
the alveoli are linear and bent in several directions.
Fig. 20. No. 3 eye-piece; 3 objective. Invasion of the large peripheral
ner\'e by the tumor.
Plate XI.
Fig. 21. No. 3 eye-piece; 6 objective. Lymphatic metastasis undergoing
cystic degeneration and showing papillary outgrowths in the wall
Fig. 22. No. 3 eye-piece; 6 objective. Invasion of a lymphatic vessel by
the polymorphonuclear cells of the tumor.
Fig. 23. No. 3 eye-piece; 6 objective. Typical soft adeno-carcinomatous
form of the tumor.
Plate XII.
Fig. 24. No. 3 eye-piece; 6 objective. The same stained with Mallor/s
aniline-blue to show the relation between the stroma and epithelial cells.
Plate XIII.
Photograph i. (Rat 47, p. 23.) Large, omental nodule, multiple, small
omental and mesenteric nodules, and nodules in the diaphragm, all developing
from a single fragment introduced into the peritoneal cavity.
Digitized by
Google
Simon Flexner and J, W. Jobling. 51
Plate XIV.
Photograph 2. (Rat 46, p. 24.) The tumor growing about and into the
cardia of the stomach where a cup-shaped ulcer is produced.
Photograph 3. (Rat 231, p. 26.) A large tumor metastasis in the axillary
gland and smaller metastasis in the retroperitoneal and other lymphatic glands,
and a g^rowth into the pleural cavity.
Plate XV.
Photograph 4. (Rat ^31, p. 26.) Metastasis in the axillary and retroperi-
toneal glands, and in the lungs.
Photograph 5. (Rat 424, p. 27.) The development of multiple nodules
from an intraperitoneal injection of ascitic fluid containing microscopic tumor
fragments.
Plate XVI.
Photograph 6. (Rat 2^3, p. 42.) The histological appearance of the typical,
soft, pure, adenomatous and acinous growth.
Digitized by
Google
THE BIOLOGY OF A MIXED TUMOR OF THE RAT.
By J. W. JOBLING, M.D.
(From the Laboratories of the Rockefeller Institute for Medical Research,
New York.)
The tumor, of which the pathological anatomy has just been fully
described, has been studied under a number of artificial conditions
with a view of establishing certain facts regarding its biology. The
more extensive biological studies of tumors in the lower animals
made up to the present time have been carried out on mice, and to a
much less extent on dogs and rats. Indeed, there does not exist in
the literature such an exhaustive study of a tumor in other animals
comparable to many studies which have been carried out in connec-
tion with the transplantable tumors of mice. It is, therefore, of
some interest to ascertain to what extent the general facts estab-
lished for tumors of mice are especially true for the transplantable
tumors of other species of animals, so that in the description to
follow, along with certain observations applying peculiarly to this
tumor, others will be described which are based on previous experi-
ments of a similar nature made upon mice.
GROWTH AND RETROGRESSION.
The phenomena of growth of the tumor were studied under a
number of artificial conditions. The first of these which we are to
consider refers to the transplantation generations, and particularly
to the percentages of successful implantations and the percentages
of retrogressing tumors in the several series. We will discuss these
points in relation to the first twenty-five transplantation generations.
Beginning with the third generation, the number of rats inocu-
lated for each one ranged from 13 to 125. The lowest per cent, of
successful inoculations w^as observed in the third generation, since
only four animals of the thirteen inoculated developed tumors, and
of these four, two finally disappeared by retrogression. The age of
52
Digitized by
Google
J. W. Jolling, 53
the tumor used for the inoculations was 132 days, and as no careful
selection was made of the fragments transplanted, this factor doubt-
less played a part in the result. The tumor employed for inocula-
tion in the fourth generation was forty-eight days old and measured
one centimeter in diameter. Of the 35 rats inoculated, 24, or 68.5
per cent., developed tumors, of which 4, or 16.6 per cent., later
underwent retrogression. From this time on the tumors chosen for
use in inoculating the other rats were about one centimeter in diam-
eter, although they were of very different ages. The youngest
tumor employed was 23 days old, and the oldest, 71. In the seventh
generation the number of successful inoculations reached 96.6 per
cent. That is, of 30 rats inoculated, 29 developed tumors, of which
6, or 20.7 per cent., subsequently disappeared. From this time on,
in spite of certain fluctuations, the percentage of successful inocula-
tions remained high, being usually 90 per cent, or over, and falling
once only, namely, in the twenty-third generation, to 68.4 per cent.
In the twentieth generation all of the 29 rats inoculated developed
tumors.
Along with the development of increased power of growth in the
inoculated rats, there went a diminished tendency to spontaneous
absorption of the tumors. Until the twelfth generation the number
of retrogressions was high and reached sometimes 52.7 per cent, in
the tenth generation, and 32 per cent, in the twelfth. After this,
the percentage of retrogressions tended to remain below 10 per cent. ;
once it fell to 2 per cent., and once, namely, in the twenty-second
generation, it rose as high as 23.6 per cent.
That the tumor is a slowly growing one is readily seen from the
length of time the inoculated animals survived the implantations
and the relatively small size which the tumors attained at the time
of the death of the rats. The tumor causes death probably in one
of three ways : first, through ulceration of the skin and secondary
infection ; second, through invasion of vital parts ; and third, through
metastasis. In respect to the degree of local development, the tumor
stands far behind the Jensen and Lewin rat tumors, which we have
also had the opportunity of studying. After the implantation of
fragments the earliest evidences of increase in size were obtained on
the seventh to the tenth day. From that period the growth is pro-
Digitized by
Google
54
Biology of a Mixed Tumor of the Eat.
gressive and quite uniform, but not rapid, and the tumors reach
sometimes a considerable size, measuring not infrequently three
centimeters in their greatest diameter.
ACTIVE IMMUNITY.
It is now established, chiefly as a result of the accurate studies of
Gaylord and Clowes,* Jensen, Ehrlich,^ Bash ford,* and still others,
that mice which first developed transplantation tumors, subse-
quently undergoing absorption possess a high degree of refractori-
ness, comparable perhaps to a state of active immunity to subsequent
inoculations of similar or even diverse tumor fragments. We
observed in conformity with this established fact that the rats, which,
having developed tumors, subsequently lost them through absorp-
tion, showed a considerable degree of refractoriness upon reinocu-
lation with the same kind of tumor fragments. An experiment was
then made to determine whether this state of resistance or immunity,
so-called, was a fixed or enduring quality or whether it was subject
to quantitative changes produced through lapse of time. The fol-
lowing tabulation bears on this question :
TABLE I.
Influence of Time Period on Reinoculability of "Immune" Rats.
Elapsed period since disap-
pearance.
Number rdnocnlated.
tumors.
Number in which
tumors disappeared.
First 30 days
30 to 60 days
60 to 90 days
90 to 150 days
26
4
I
5
4
I
Totals
61
10
I
Control
80
69
21
Table I shows, first, that the state of refractoriness left by the
retrogression of growing tumors is not perfect, but is of high degree;
and it further indicates that this degree is greatest immediately after
the disappearance of the tumors, and becomes reduced subsequently
* Gaylord and Clowes, Med. News, 1905, Ixxxvii, 698.
'Ehrlich, Arb. a. d. k. Inst. f. exper. Therapie, 1906, No. i, 77.
•Bashford, Proc. of the Royal Sac, Series B, 1907, Ixxix, 164.
Digitized by
Google
J. W. Jobling.
55
by mere lapse of time. Within a period of 90 days after the retro-
gression of the timior in 61 rats, 10, or 16.5 per cent., proved rein-
oculable with a tumor which yielded in the control animals 86.2 per
cent, of tumors. Moreover, of the 10 tumors developing in the
refractory rats, only one was afterwards absorbed, and that was in
a rat coming within the first thirty-day period since the original
absorption. Among the control rats, on the other hand, the retro-
gressions were considerably higher, and equalled 31.8 per cent.
It is significant that on a second attempt to implant successfully
tumor grafts in 13 rats that had resisted a first reinoculation, from
30 to 150 days previously, none were successful. The control rats
of this series gave 75 per cent, of successful implantations.
NATURAL REFRACTORINESS.
It has been repeatedly observed by all who have studied trans-
plantable tumors that animals which resist implantation of tumor
grafts are not wholly immune, but are merely refractory, and can
often be successfully inoculated with a similar tumor of higher viru-
lence, or, as is sometimes stated, with greater capacity for growth.
We were interested not only in confirming this result with the rat
tumor, but also in ascertaining to what extent the effect of the first
unsuccessful implantation tended to add to the natural refractory
state of the animals. Having ascertained that mere lapse of time
was attended by diminution of this refractory state in some rats
(Table I), a test similar to that employed in them was now applied
to the rats which resisted a first inoculation. The result is shown
in Table II.
TABLE IL
Influence of Time Period on Inoculability of Negative Rats.
Elapsed pariod since
Number reinocalated.
Number and per cent,
developing tumors.
Number in which
tumors disappeared.
First 30 days
30 to 60 days
60 to 90 days
90 to 120 days
120 to 150 days
17
60
62
47
14
2(11.7)
31 (Si-6)
33 (S3»)
»7 (S7.4)
5 (3S-7)
12
3
3
I
Totals
200
98(49)
19
Controls
80
69 (86.a)
21
Digitized by
Google
56
Biology of a Mixed Tumor of the Bat.
But before discussing the results, another tabulation (Table III)
will be given, since it expresses the effect of a third implantation of
tumor fragments in rats which had resisted two previous inoculations.
TABLE III.
Influence of Time Period on Inoculahility of Negative Rats.
Elapsed period since
second negative inoculation.
Number reinoculated.
Number and per cent.
Number in which
tumors disappeared.
First 30 days
30 to 60 days
18
17
a(ii.i)
3 07.6)
I
ToUl
35 1 5 (14.2)
I
Control
20 1 15 (75)
6
These tabulations indicate clearly that failure in respect to the
tumor implantati >ns depends upon factors which are not controlled
entirely by the animal subjected to the inoculation, or even by the
fragment of tumor implanted, but represents the joint result of the
action of both sets of conditions on each other, namely, those inher-
ent in the animal and those inherent in the tumor. The virulence
or the capacity for growth of the tumor used to inoculate the nega-
tive animals at the time of the first inoculation did not exceed the
average of the tumors employed for implantation at this period
of the study, and yet 49 per cent, of successful inoculations were
secured. And while the tumor used to inoculate the negative
animals at the time of the second inoculations showed among
the controls a lower virulence or a lessened capacity for growth
than the tumor used for the first reinoculation, yet a further 14.2
per cent, of successful implantations were secured.
The case is not so clear as regards the natural refractory state's
being increased by a first unsuccessful inoculation. Table II is sug-
gestive of such an effect as regards the small number of successful
reinoculations in the first thirty-day period following the unsuc-
cessful implantation, as compared with the greater number of suc-
cessful inoculations at later periods. But the figures given are
indications merely, and do not serve to establish the point.
Digitized by
Google
J. W. Jdbling.
57
SECONDARY AND TERTIARY IMPLANTATIONS.
Ehrlich* has pointed out that in the case of actively growing
mouse tumors, secondary implantations undertaken at a time at
which the first tumor is growing rapidly usually fail to grow. With-
out discussing the theoretical views which he has offered to explain
this observation, we wish to report the results of secondary and
tertiary implantations carried out with this slow growing rat tumor.
As has already been pointed out, this tumor exhibits in a high degree
the property of metastasising in nearby and distant parts of the body,
from which it is concluded that the effect on the organism of a
growing tumor was not such as to interfere with or inhibit the
growth of a secondary tumor. Table R^ exhibits the results of the
reinoculations, and serves to bring out several points of consider-
able interest.
TABLE IV.
Reinoculation of Rats with Growing Tumors.
First 30 days
30 to 60 days
60 to 90 days
90 to 120 days
120 to 150 days
Control
0.7
I.I
"3
1.3
V
a
Iriii
63
IS
5
36
43
78
II
5
I "8
So
bo
3 S I
69
rill
36
23
62
9
2
z II
21
fall
cog
27
10
5
4
o
E4
It shows, in the first place, that a growing tumor does not prevent
the successful implantation of a second tumor of the same kind, but
it also indicates that the secondary implantations are likely to yield
a smaller number of growing tumors, the susceptibility of the rats
inoculated having been established, than the primary inoculation.
The influence of the time element during which the first tumor has
* Ehrlich, Arb, a. d. k. Inst. f. exper. Therapie, 1906, No. i, 77.
Digitized by
Google
58 Biology of a Mixed Tumor of the Bat.
been growing in the body appears to be of some importance in deter-
mining the result, but the factor that exercises a definite and deter-
mining influence is not the original susceptibility of the animals to
inoculation, but the state of the primary tumor with reference to
its later history. So long as the primary tumor is itself expanding
the secondary inoculations are readily accomplished, but when the
primary tumor has become stationary in size, or has begun to undergo
retrogression, the secondary implantation usually fails. In only
one instance among 46 cases did a rat in which the primary tumor
was being absorbed develop a secondary tumor. In view of the
fact that rats showing retrogressing tumors do not develop sec-
ondary tumors, while those with growing tumors tend to develop
them, it would be expected that few or none of the secondary tumors
should later suffer retrogression. Table IV shows that in only two
rats was there a disappearance of the secondary growths.
The influence of the period of growth of the primary tumor on
the result of the secondary implantations is shown by the fact that
in the first thirty-day period all the rats in which the primary tiunor
continued to grow developed secondary tumors, and that some rats
with growing tumors in the later periods did not develop secondary
tumors.*^
Still another test of the reinoculability of rats with growing
tumors w-as carried out with 29 animals in which two tumors w^ere
growing at the time of the third inoculation. This implantation
was made during the period of from 30 to 60 days of the duration
of the two growing tumors, with the result that of the 29 rats inocu-
lated 23 developed tertiary tumors. It follows from these observa-
tions that with the rat tumor with which these tests are made
reinoculation is possible in all or nearly all the animals, provided
the tumors which originally developed from the first implantation
are still growing, and have not either become stationary, or begun
to undergo retrogression.
" Recently Gay has attempted to show that rats inoculated with this tumor are
subject to reinoculation only after the period of metastasis has been reached,
and that at an earlier period the result of a secondary inoculation is to cause
rather absorption of the original tumor than the successful implantation of the
second. These statements are in direct conflict with the observations repre-
sented by Table IV.
Digitized by
Google
J. W. Jobling.
59
PROMOTING INFLUENCE OF TUMOR EMULSION.
In the course of experiments performed with a number of organic
and other substances to be related, the influence on the growth of
the tumors of an emulsion of the tumor substance itself was studied.
Previous experiments had shown that fine emulsions of the tumor
TABLE V.
Effect on Tumor Growth of Heated and Unheated Tumor Emulsion.
Treated with emulsion.
Control.
Mode of treatment.
^1
Number and
per cent.
developing
tumors.
Number and
per cent, in
which tumots
disappeared.
If!
i|l
Unheated emulsion one day and tumor
next day
Heated emulsion one day and tumor
next day
8
10
7 (87.5)
10 (ico)
3 (42.7)
5(50)
100
20
70
50
Tumor one day and heated emulsion
next day
Unheated emulsion subcutaneously and
tumor lo days after
9
10
9(100)
6(66)
6(66)
6(100)
100
80
70
75
Unheated emulsion intraperitoneally
and tumor lo days after
Heated emulsion subcutaneously and
tumor lo days after
10
28
7(70)
28(100)
6(85)
4(14)
80
82.5
75
36
Heated emulsion twice at lo-day inter-
vals and tumor lo days after second
injection
Heated emulsion three times at lo-day
intervals and tumor lo days after
third injection
10
10
10 (100)
10 (100)
1(10)
70
87
56.2
83.6
Heated emulsion and tumor 20 days
after
Heated emulsion and tumor 30 days
after
10
10
10 (100)
10 (100)
2(20) .
4(40)
70
87.5
56.2
83.6
Heated emulsion twice at lo-day inter-
vals and tumor 20 days after second
injection
Heated emulsion three times at lo-day
intervals and tumor 30 days after
third injection
10
8(80)
87.5
83.6
9
9(100)
i(n)
90
55.5
in salt solution did not give rise to tumor formation, so that there
was little or no risk of having tumors develop from the injected
emulsion. The emulsion which was made of a uniform milky
appearance was divided into two portions, one being subjected to
Digitized by
Google
60 Biology of a Mixed Tumor of the Bat.
heat in a water-bath, at a temperature of 55° C. for thirty minutes
before the injection, and the other not being further treated. A
standard suspension of lecithin was used as control. Injections of
one cubic centimeter of the emulsion were made into the peritoneal
cavity and the tumor material used was of the type of growth not
definitely adeno-carcinomatous. The experiments have not been
repeated since the tumor assimied the latter form. At a later period,
and, therefore, subsequent to the injection of the emulsion, tumor
fragments were inoculated beneath the skin, and the results observed,
as recorded in Table V.
As the table shows, the effects of the injections of the emulsion
are considerable. They will now be taken up separately for con-
sideration.
The effects of the unheated and of the heated emulsions begin to
be displayed as early as twenty- four hours after their injection, and
in about the same degree. Thus the nimiber of positive inocula-
tions is the same in both series, and the number of retrogressions
is also about the same. As compared with the control series, the
effect of the unheated emulsion is seen in the smaller number of
disappearing tumors, and of the heated emulsion in an increasing
number of positive implantations. It chanced that the tumor chosen
for the experiments with the heated emulsion gave in the controls
a low percentage of positive implantations, but as the same tumor
was inoculated into the animals previously having had injections of
the heated emulsion, the contrast is very marked. On the other
hand, the injection of the heated emulsion on the day following the
implantation of the tumor morsels is followed by little or no special
influence on the number of retrogressing tumors, since they were
about the same as in the control series.
The next set of experiments yielded much more striking results.
Unheated emulsion was injected subcutaneously and intra-perito-
neally, and heated emulsion subcutaneously ten days before the
tumor implantations. The rats which had received the unheated
emulsion behaved very much as did the controls, since the number
of positive inoculations and of retrogressions was about average,
while the rats receiving the heated emulsion gave 100 per cent,
of positive implantations, and a small number, 14 per cent., of
Digitized by
Google
J. TT. Jdbling.
61
retrogressions, as against 82.5 per cent, of positive implantations
and 36 per cent, of retrogressions in the control series. It appears,
then, that rats treated ten days in advance with a heated tumor
emulsion show increased susceptibility to the tumor implanta-
tions and a diminished tendency to tximor retrogressions, as com-
pared with untreated rats or rats treated with unheated emulsion.
Indeed, the unheated emulsion proved not to exhibit any predispos-
ing effect whatever on the growth or the persistence of the tumors.
Another series of experiments was performed with the heated
emulsion. The experiments were so arranged that the injections
were made two or three times at ten day intervals, and the tumor
implantations were conducted at ten, twenty, and thirty day inter-
vals after the second and third injections. These experiments will
show the promoting influence of the heated emulsion on the growth
and endurance of the tumors. We possess no knowledge of the
nature of the constituent of the heated emulsion upon which depends
the effect described, or of the mechanism of the promoting influence.
In view of facts already established for mice tumors, that blood cor-
puscles and several kinds of tissues of the same animal species exer-
cise, when injected in advance of the tumor implantations, a restrain-
ing influence on the tumor growth, the absence of such a restraining
influence following the development of the unheated emulsion of
txmior and the development of the promoting qualities following the
injection of the heated emulsion are points worthy of special atten-
tion. It remains to be ascertained whether this effect is peculiar
or limited to this rat tumor, or to a small number of transplantable
tumors, or whether it is applicable equally or in some degree to
transplantable tumors in general. We are engaged at present in
elucidating this point.
TABLE VI.
Summary of Effect on Growth and Persistence of Tumor of Heated Emulsion.
Series.
Total number of rats
treated.
Number and per cent,
developing tumors.
Number and per cent.
in which tumors
disappeared.
Control
Unheated emalsion
Heated emulsion
28
28
106
26 (92.8)
20 (74.2)
104 (98.1)
19 (73)
15 (75)
23 (22)
Digitized by
Google
62 Biology of a Mixed Tumor of the Bat.
We have summarized in Table VI the effects of the heated emul-
sion on the growth and persistence of the tumor.
Thus it will be seen that among 28 control rats 26 developed
tumors, equalling 92.8 per cent, of which 19, or 73 per cent., later
underwent retrogression. Among 28 rats treated with unheated
emulsion, 20, or 74.2 per cent, developed tumors, of which 15, or
75 per cent., underwent retrogression. These figures are to be
contrasted with the next, in which of 106 rats treated with the heated
emulsion, 104, or 98.1 per cent., developed tumors, of which 23, or
22 per cent., later were absorbed.
The fact should be mentioned that when a very large number of
animals were to be inoculated a large tumor was always selected,
so as to insure uniform results. As the large tumors are more
advanced in degree than the small ones, they tend to give a lower
percentage of successful implantations and a higher percentage of
retrogressions. In the present case these points emphasize the
effects of the heated emulsion.
EFFECT OF HEATED EMULSION ON IMMUNE RATS.
As has been already pointed out, rats which have recovered spon-
taneously from the tumors possess an increased degree of resistance
to reimplantation of tumor fragments. In view of the power of
the heated emulsion to overcome the natural resistance to the implan-
tation of the tumor, it became desirable to ascertain whether this
induced or increased resistance could also be set aside by the emul-
sion. The experiments made to test this point were carried out on
two small series of rats which had recovered from tumors that had
developed in animals previously receiving the heated emulsion.
In the first series of eight rats, the tumors had entirely disap-
peared for a period of from 30 to 90 days, and they received no new
injection of the heated emulsion. Three of the eight rats developed
tumors from the new implantations, of which one tumor subse-
quently disappeared. The second series of nine rats had been free
of tumor for a period of from 30 to 60 days. The heated emulsion
was again injected ten days prior to the second tumor implantations.
Four of these developed tumors, of which two subsequently disap-
peared. The control rats for these series gave 100 per cent, of
Digitized by
Google
J. W. JoUing.
63
successful inoculations and no retrogressions, indicating that the
tumor used for implantation was of maximum virulence. There
is, therefore, no evidence, so far as this experiment goes, of any
influence of the emulsion in promoting successful implantation in
these rats.
The experiment was now extended to rats which had not been
previously injected with the heated emulsion, but which had recov-
ered spontaneously from tumors. They received the heated emulsion
in the usual way, and afterwards implantation of tumor fragments.
Two series of control animals were employed: one consisting of
animals not previously inoculated, to establish the virulence of the
tumor, and the other " immune " rats, so-called, in which no emul-
sion had been introduced. Table VII summarizes this experiment,
from which it will be seen that of ten rats in which timiors had dis-
appeared, from 30 to 120 days previously, one only developed a
tumor, and this subsequently disappeared ; whereas the ten control
animals, which had not been previously inoculated, all developed
TABLE VII.
Ejfect of Heated Emulsion on Immune Rats. No Emulsion (control.)
Time period since disappear-
ance of tumor.
Number inoculated.
Number developing
tumors.
Number in which
tumors disappeared.
30 to 120 days
10
I
I
Control
ID
10
Emu/sum on Same Day as Tumor,
30 to 90 days
Emulsion Ten Days Before Tumor.
30 to 120 days
tumors, of which none disappeared. These results are to be com-
pared with those obtained in animals in which the injection of the
emulsion and the implantation of the tumor were made on the same
day, and the emulsion injected one day and the tumor implanted ten
days subsequently, as are shown in the lower part of the same table.
Thus, of ten rats which had recovered from 30 to 90 days pre-
viously, and which received the emulsion and the tumor on the same
Digitized by
Google
64 Biology of a Mixed Tumor of the Bat.
day, three developed tiimors ; whereas of the ten which had recov-
ered from 30 to 120 days previously, and which received the tumor
ten days after the heated emulsion, five developed tumors, in both
cases there being no retrogressions.
The experiments are quite suggestive and indicate that the heated
emulsion can overcome in some degree the induced resistance or
immunity produced by the absorption of the tumor, and also that
while this influence exerted by the emulsion is greater in rats which
had not previously received the emulsion, it may still be exerted
upon animals possessing the increased resistance given by recovery
from the tumor following the emulsion. Moreover, the effect begins
to be apparent on the same day as the injection of the emulsion,
although it is more pronounced ten days afterwards.
EFFECTS OF THE HEATED EMULSION ON RATS WITH DISAPPEARING
TUMORS.
Rats with growing tumors may be inoculated a second and a third
time with tumor grafts. Rats recovering spontaneously from tumors
acquire an increased resistance to tumor implantation. Hence,
recovery from the tumor is attended by a rise in resistance in the
organism associated with the absorption of the tumor, or to local
conditions that affect the nutrition of the growth. The interesting
question, therefore, arises as to whether this state of increased
resistance which prevails during the retrogression is general or local,
and whether it is capable of preventing the development of a secon-
dary graft and can be overcome by the promoting influence on the
tumor growth exercised by the heated emulsion. The incidental
question was also put, namely, whether a secondary graft in an
animal with a disappearing tumor can affect that tumor in such a
way as to bring about a renewal of its growth. Table VIII supplies
the answer to these questions as far as they have been worked out.
By a system of measurement we were able to ascertain fairly pre-
cisely the period at which growth ceasecj in the tumors and retro-
gression set in. In this experiment particular attention was paid
to these two points. The results seem to indicate that the heated
emulsion may overcome in some degree the condition of the organism
Digitized by
Google
J. W. Jobling.
66
that operates to bring about tumor absorption. This indication,
which we wish to state guardedly, is based on the observation that
of the ten rats with retrogressing tumors receiving the emulsion,
in only two did the tumors progress to complete disappearance,
which is in contrast to the usual observation. Generally speaking,
TABLE VIII.
Secondary Inoculation of Rats with Disappearing Tumors. No Treatment.
30 to 90 days
G>IltTOl
s
c .
B
^i
a
4I
m
all
in
-If
All developed tumors
"sip
1.1^
••s
s-Sg§
•s s. •
® S 2
Emulsion Ten Days Before Tumor.
30 to 90 days
10
3 I 1 4
4 1 2
I
Control
10
All developed tumors
1
when retrogression has once begun, it proceeds to complete disap-
pearance, provided the animals survive the necessary period. More-
over, while in the control series no new implantations developed in
animals in which the primary tumors disappeared, yet one of three
such animals of the series receiving the emulsion developed a sec-
ondary tumor.
A further result of this experiment was the successful implanta-
tion of new tumor grafts upon rats with retrogressing tumors. In
view of this result, it may be considered as doubtful whether the
state of the organism as a whole influences at all times the local
tumors in respect to growth and recession, and whether local activi-
ties, and particularly the nutritive conditions, may not well play at
times the determining part. It has been observed by Velich, Loeb®
and others that the excision of a part of a stationary or retrogressing
timior nodule may be followed by renewed activity, which is expli-
cable most readily on the supposition that the limitation of growth
is affected by local conditions. But a somewhat more significant
point developed by this experiment is the awakening, as it were, of
•Loeb, Arch. f. Anat., Physiol, u. wissensch. Med., 1902, clxvii, 175.
Digitized by
Google
66 Biology of a Mixed Tumor of the Rat
growth in the receding tumors by means of a secondary inoculation.
Thus in seven of twenty rats showing receding tumors, a new growth
began, following and apparently as the result of the reinoculation of
the secondary graft.
EFFECTS OF ORGANIC SUBSTANCES ON TUMOR GROWTH.
The question which next arises concerns the element of the tumor
emulsion upon which its promoting influence depends. We have
already stated that we have not been able to determine this sub-
stance. We have, however, endeavored to ascertain whether the
substance resides in the fluid serum or lymph contained within the
emulsified tumor, or within the cells themselves. We have ap-
proached this question indirectly by testing the effect of the blood
serum of the rat itself. We have, moreover, had in mind in the
planning of these experiments the statement of Gaylord and
Clowes,'' that the blood serum of mice which have recovered spon-
taneously from implanted tumors exercises an immunizing or re-
straining effect on the development of implanted grafts of mouse
tumors, and these effects are not exhibited by the serum of normal
rats. It is true that Michaelis,® Ehrlich® and others have not con-
firmed this experiment, but since it lay in our way, we have not
failed to make certain observations to cover this point.
For the purpose of this experiment the blood serum from several
different series of rats was employed. Thus, the effects of the
serum obtained from normal rats was compared with those of the
serum obtained from rats which had failed to develop tumors on
inoculation, and, hence, were designated "negative"; from rats
which had recovered spontaneously from growing tumors, and
designated "immune"; and finally, from rats possessing growing
tumors at the time the serum was collected, and designated " posi-
tive." This series of experiments was controlled with bouillon.
Ringer's solution, normal horse serum, and heated and unheated
emulsion of the tumor itself. The procedures were in different
experiments somewhat varied, so that this serum and the other sub-
' Gaylord and Clowes, Johns Hopkins Hasp. Bull, 1905, xvi, 130.
• Michaclis, Zeit. f, Krebsforsch., 1906, iv, i.
• Ehrlich, Arb. a, d. k, Inst, f, exper, Therapie, 1906, No. i, 77.
Digitized by
Google
J. W. Johling.
67
stances were given sometimes before and sometimes after the
tumor implantations, as is indicated in the tables.
Table IX summarizes the experiment made with these substances
which were given on the day preceding that on which the tumor
implantations were made. *
TABLE IX.
Effect on Tumor Implantations of Blood Serum, etc., Given Preceding Day.
Substance injected.
Number of rats
inoculated with tumor.
Number and per cent,
developing tumors.
Number and per cent.
in which tumors
disappeared.
None— control
Bouillon — control
Normal rat semm
"Positive" rat serum
"Negative" rat serum
** Immune " rat serum
lO
lO
10
i
4
10 (100)
4(ioo)
3(30)
4(80)
7 (87.5)
3(60)
2(40)
1(25)
The stimmary is instructive, since it shows, apparently, that the
bouillon exercises a certain restraining effect, and that next to the
bouillon the normal rat serum is the most influential substance. On
the other hand, the serum from the so-called " positive " rats did not
restrain the original taking of the grafts, while of the series of five
animals employed three subsequently suffered retrogression of the
tumors, which is a much higher percentage than the retrogressions
suffered by the control series. The serum from the rats designated
" negative " and " immune " affected the original implantations and
retrogressions practically not at all as compared with the control
series.
TABLE X.
Effect on Tumor Implantations of Blood Serum, etc.. Given Following Day.
Substance injected.
Number of rats inocu-
Number and per cent.
Number and per cent, in
lated with tumor.
developing tumors.
which tumors disappeared
None — Control
10
a (20)
I (50)
Bouillon— Control
10
3 (30)
2 (66)
Horse serum — Control
10
2 (ao)
» (so)
Ringer's solution — Control
10
6 (60)
5 83
Normal rat serum
10
6 (60)
5 (83)
" Positive " rat serum
410
3 (75)
3 iioo)
"Negative" rat serum
10
4(40)
2 150)'
** Immune " rat serum
10
4 (40)
10 (100)
3 <7S)
Heated emulsion
10
5 (50)
"Six of the ten rats inoculated died soon after the inoculation.
Digitized by
Google
68
Biology of a Mixed Tumor of the Bat
Table X exhibits the effects of the injections of the various sub-
stances carried out on the day following the tumor implantations.
It chanced that in this experiment the implanted tumor was of
poor growing property. In some respects this was an advantage,
inasmuch as it allowed for the play and exhibition of influences that
are not so easily detectable where the power of growth is maximum.
It is clear from the experiment that the rat sera tend rather to pro-
mote than to inhibit the success of the implantations, but what is par-
ticularly striking in this experiment is the promoting influence on
the original tumor graft of the heated emulsion.
Table XI exhibits the influence of these substances when given
ten days before the tumor implantation.
TABLE XI.
Effect on Tumor Implantation of Blood Serum, etc.. Given Ten Days Before.
Substance injected.
Number of rats in-
Number and per cent.
Per cent, in which
oculated with tumor.
developing tumors.
tumors disappeared.
None — Control
ID
8 (8o)
75
Bouillon— Control
lO
7 (70)
6 (60)
71
Horse serum — Control
lO
33
Ringer's solution— Control
lO
7 (70)
71
Normal rat serum
lO
9 (90)
33
"Positive" rat serum
10
9 (90)
33
"Negative" rat serum
lO
8 (80)
62
*< Immune " rat serum
10
6 (60)
33
Unheated emulsion, subcutaneously
lO
100
Unheated emulsion, intraperitoneally
lO
7 (70)
85.6
Heated emulsion intraperitoneally
9
9 (loo)
22
It supports the previous observations in indicating that the nature
of the rat serum plays no part in determining the result of the
implantations, and that of all the substances employed the heated
emulsion alone produces a marked and undeniable influence on the
growth and persistence of the grafts.
The several tables bring out in spite of certain irregularities the
important data sought by the experiments. It is perhaps possible
to inhibit in some degree tumor development by injecting into the
body a short time before the tumor fragments are implanted certain
relatively indifferent fluids, of which bouillon seems to be as active
as any. On the other hand, the injection at any period, as regards
the tumor implantations, of the serum of the blood of the rat
Digitized by
Google
J, W. Jdbling.
69
whether obtained from normal animals, animals carrying tumors,
failing to develop tumors, or recovering from them, entirely fails
to produce any effect either in the direction of restraining or of
promoting growth. It is, therefore, safe to conclude that the prop-
erty affording resistance or relative immunity to certain rats to the
primary inoculation, or developed through spontaneous recovery
from tumors once successfully implanted, does not reside in the
serum of the blood of these animals.
EFFECT OF BLOOD CORPUSCLES AND OTHER TISSUES ON
TUMOR GROWTH.
Bashford" observed that the injection of the blood of the mouse
exercised a restraining influence on the development of mouse
tumors. Schoene^^ observed that the injections of suspensions of
mouse embryos produced a similar effect, and Michaelis*^ and
BorreP* observed that the injections of suspensions of liver and
spleen acted apparently in the same manner. The activity of the
blood in this respect was shown by Bash ford** to depend on the
corpuscles, and not to reside in the serum, and it may be assumed
that in the other cases mentioned it is also the cellular elements and
not the fluids upon which the activities depend. The experiments
to be related refer to the effects of blood corpuscles separated from
TABLE XII.
Effect on Tumor Growth of Blood Corpuscles Given on Preceding Day.
Substance injected.
Number of rata
inoculated with tumor.
Number and per cent,
developing tumors.
Number and per cent.
in which tumors
disappeared
None— control
Bouillon — control
Ringer's solution— control
Normal rat serum
Washed rat corpuscles
Unheated emulsion
Heated emulsion
lO
10
lO
9
10
8
9
10 (lOO)
6(6o)
9 (90)
8(88}
9(100)
7(7o)
liSi
I (I2.S)
4 (66.6)
3 (42-8
6 (66.6)
" Bashford, Annual Report of the Imperial Cancer Research Fund, 1906, iv, 5.
""Schocne, Munchener med. Woch., 1907, liii, 2517.
"Michaelis, Zeit. f, Krebsforsch., 1906, iv, i; Deut. med. Woch., 1907,
xxxiii, 1826.
" Borrcl, Bull, de Vlnst. Pasteur, 1907, v, 605.
Digitized by
Google
70
Biology of a Mixed Tumor of the Bat.
the serum, washed, suspended in salt solution, and injected into the
body, on the tumor implantations. The blood corpuscles and the
substances used for controls were injected on the day preceding the
implantation of the tumor, or ten or twenty days preceding that
operation.
Table XII exhibits the effects produced on the tumor growth by
an injection of corpuscles made on the day preceding the tumor
grafts.
At the end of this brief period no effect has been produced, since
the number of successful inoculations is about the same as those of
the controls, which indeed is maximum, while the number of retro-
gressions is about the equivalent of the controls, indicating that no
marked inhibition of growth through the development of increased
resistance has been produced. On the other hand, the effect of
the injections of blood corpuscles ten days before the timior implan-
tation is unmistakable, since, as Table XIII shows, not only is the
number of successful inoculations greatly reduced, as compared
with the control, but all of the tumors which originally developed
underwent subsequent retrogression and disappeared.
TABLE XIII.
Effect of Blood Corpuscles Given Ten Days Before Tumor Implantation.
Substance ii^eaed.
Number of rats
Number and per cent,
developing tumors.
Number and per cent.
in which tumors
disappeared.
None*— control
Normal rat serum
Washed corpuscles
10
10
lO
10 (lOO)
8(8o)
4 (40)
4 (So)^
4(100)
The effect of the blood corpuscles is still present as late as twenty
days after their injection, as can be seen by reference to Table XIV.
TABLE XIV.
Effect of Blood Corpuscles Given Twenty Days Before Tumor Implantation,
Subcance ii^ected.
Number of rats
inoculated with tumor.
Number and per cent.
Number and per cent.
in which tumors
disappeared.
None— control
Normal rat serum
Washed corpuscles
ID
10
10
10 Tioo)
lo (100)
7(70)
3(33)
a (201
5 (71.4)
Digitized by
Google
J. W. Jobling.
71
Apparently the effect is disappearing, since the number of suc-
cessful implantations has not only increased over those of the ten-
day period, although still smaller than the controls, but the number
of retrogressions has also diminished.
The effects, therefore, of the injection of washed corpuscles of
the rat into the peritoneal cavity on the development of grafts of
this rat timior placed beneath the skin is shown by these experiments
to be considerable. The effect is not developed within twenty-four
hours of the injection of the corpuscles, at which time such inhibi-
tion as has been afforded by bouillon, for example, is at its height.
But at the expiration of ten days the inhibitory effect of the washed
corpuscles is very great, and this effect is still appreciable at the
expiration of twenty days, but how much longer it persists we have
not undertaken to ascertain. This is in conformity with Bashford's
observations on mice, who found that the refractory condition
produced by the injection was not so marked four days after the
injection as at the expiration of ten days, and that it persisted for
at least three weeks.
The final experiment of this series was made with emulsions of
several organs in a manner similar to that employed with the blood
corpuscles and also with the emulsions of the tumor. These sus-
pensions were injected into the peritoneal cavity, and ten days later
the implantations of the tumor fragments were made beneath the
skin. Table XV summarizes the results of this experiment.
TABLE XV.
Effect on Tumor Growth of Organ Emulsions Given Ten Days Before,
Substance injected.
Number of rats
inoculated with tumor.
developing tumors.
Number and per cent.
in which tumors
disappeared.
None — control
lO
8(80)
6(60)
7(75)
3 (50)
Unheated liver
lO
Heated liver
lO
5 J""!
8 (80)
I (25)
Unheated muscle
lO
3(37)
Heated muscle
lO
8(80)
8(80)
1(10)
3(37)
Unheated spleen
lO
Heated spleen
Unheat(^kidney
lO
lO
8 80)
8(80)
2(40)
3(37)
Heated kidney
lO
4 (501
Unheated testicle
lO
2 (251
Heated testicle
lO
8(80)
2(25)
Unheated tumor
lO
5 (50)
4(80)
Heated tumor
lO
10 (100)
8(80)
Digitized by
Google
72 Biology of a Mixed Tumor of the Bat.
It IS obvious from examination of the table that emulsions of the
organs produce no remarkable change in the organism of the rats,
through which the development of the tumor grafts is seriously
inhibited. There are, indeed, four places in the table which indicate
that inhibition of some degree had taken place. These are in con-
nection with the heated and unheated liver, the heated spleen, and
the unheated tumor emulsion. The greatest degree of inhibition
observed was in connection with the heated liver, but as compared
with the action of the blood corpuscles it may be said that the organs
containing much blood tend to be more inhibitory than those which
contain little blood, and no action would appear to be exerted by the
cells peculiar to the organs. Lastly, as far as the results obtained
from the heated organic emulsions can be compared with those
obtained from the heated tumor emulsions, they may be said to be
of the reverse character.
Digitized by
Google
EXPERIMENTS ON THE INFLUENCE OF RADIUM
BROMIDE ON A CARCINOMATOUS TUMOR
OF THE RAT.
By maud L. MENTEN.
(From the Laboratories of the Rockefeller Institute for Medical Research,
New York.)
INTRODUCTION.
A considerable number of observations have already been pub-
lished regarding the use of radiation by means of radium bromide
in the treatment of malignant tumors. One of the earliest inves-
tigators of the subject was Exner/ who employed it in the treat-
ment of epithelioma and sarcoma, and who was the first to report
on the microscopic examination of tumors so treated. He noted
particularly the overgrowth of connective tissue which preceded the
disappearance of the tumor cells, and which corresponded in amount
more or less with the reduction in size of the growth. The obser-
vations which followed Exner's publication were not always so
favorable. The greater number of reports emphasized the bene-
ficial effects of radium upon rodent ulcer, which unless it extends
over a wide surface is said readily to improve and to disappear com-
pletely under the influence of the radiations. Those who have
reported successful results are Lehmann,^ Foreau de Courmelles,*
^ Exner, Ueber die Behandlung von Oesophaguskarzinom mit Radium-strahlen,
Wiener klin, Woch., 1904, xvii, 96; Ueber die Art der Ruckbildung von Karzi-
nommetastasen unter der Einwirkung der Radiumstrahlen, ihid., p. 181; Ueber
die bisherigen Dauerresultate nach Radiumbehandlung von Karzinomen, Deut-
sche Zeit. f, Chirurgie, 1904, Ixxv, 379.
'Lehmann, Essai sur Taction th6rapeutique du Radium, Arch, gen, de mid.,
1906, i, 1301.
* Foreau de Courmelles, Un peu d'histoire de la therapeutique du radium,
Actualitis midicales, 1904, xvi, 84.
73
Digitized by
Google
74 Influence of Radium Bromide on Tumor of Rat.
Czerney,* Braunstein,^ McLeod,* Sichel/ Mclntyre,® Davidson,*
Scholz/^ Schwarz," Abbe,** and others. There is difference of
opinion as regards its merits in the treatment of epitheliomata
proper. Undoubtedly a considerable number of growths thus treated
begin to improve almost immediately, and gradually entirely dis-
appear, while others, apparently similar, continue on the contrary
to extend in spite of the treatment. Thus Lehmann, Abbe, Czerney
and Scholz report cures in many cases, and Foreau de Courmelles
speaks of improvement through its use, while others report either
negative or very transitory beneficial effects. The results with the
more malignant forms of tumors have been almost uniformly
unpromising. In the case of carcinoma of the esophagus producing
constriction, it has been possible through the use of a specially con-
structed apparatus to produce amelioration of some of the most dis-
tressing symptoms which result from the obstruction (Exner and
Einhom),** but in no case was a cure effected. Abbe states that
he observed carcinomata to disappear under the influence of radium,
but he also states that there was a marked tendency towards
recurrence.
The results obtained through the employment of radium in trans-
plantable mouse tumors have been more gratifying. Apolant**
•
* Czcraey, Zeit /. Krebsforsch., 1907, v, 27.
•Braunstcin, Uebcr die Wirkung der Radium-emanation auf bosartige Tu-
moren, Therapie d. Gegenwart, 1904, xlv, 412.
*McLeod, Further Observations on the Therapeutic Value of Radium, Brit,
Med. Jour,, 1904, i, 1366. * Sichcl, Brit. Med. Jour., 1904, i, 182.
• Mclntyre, Radium and Its Therapeutic Effects, Brit, Med. Jour., 1903, ii, 1524.
•Davidson, Radium Bromide, Brit, Med. Jour., 1904, i, 181.
"Scholz, Ueber die physiologische Wirkung der Radium-strahlen und ihre
therapeutische Vcrwendung, Deutsche med. Woch., 1904, xxx, 94.
"Schwarz, Uebcr die Wirkung der Radium-strahlen, PfiUgers Archiv, 1903,
c, 532.
"Abbe, The Subtle Power of Radium, Med, Record, 1904, Ixvi, 321; The
Specific Action of Radium as a Unique Force in Therapeutics, ibid., 1907, Ixxii,
583. Radium in Surgery, Jour, of the American Med, Assn., 1906, xlvii, 183.
"•Einhom, Radium Treatment of Cancer of the Esophagus, Jour, of the
American Med, Assn., 1905, xiv, 8; Radium Receptacles for the Stomach, Eso-
phagus, and Intestines, Med, Record, 1904, Ixv, 399.
" Apolant, Ueber die Einwirkung der Radium-strahlen auf das Karzinom der
Maiise, Deutsche med. Woch., 1904, xxx, 454; Ueber die Ruckbildung der
Maiisekarzinome unter dem Einfluss der Radium-strahlen, Deutsche med. Woch.,
1904, xxx, 1 126.
Digitized by
Google
Maiid L, Menten. 75
caused the inoculated carcinoma in mice to undergo retrogression
in all the animals exposed to the radiations, and a large percentage
subsequently entirely disappeared. He states as a result of the
microscopical examination of tumors exposed to radium that no
relation existed between the amount of radiation and the degree of
change induced, and he advanced the theory that the primary action
of the rays was on the tumor cells, and that there resulted from this
a secondary stimulus which led to an overgrowth of the connective
tissue that was characteristic of the disappearing tumors. Bash-
ford,^* who used the Jensen tumor, corroborated Apolant's results,
but he attributed the injurious action of the radium to a variety of
factors.
EXPERIMENTAL OBSERVATIONS.
The radivmi used in the experiments to be related was kindly
supplied by Mr. Hugo Lieber, of New York, and later by Dr.
Robert Abbe. The specimens supplied by Mr. Lieber were con-
tained in hard rubber capsules covered with a mica plate, and in the
form of celluloid coatings. Of the former there were three grades :
(i) 10 milligrams of so-called 1,000,000 activity; (2) 25 milli-
grams of so-called 10,000 activity; and (3) 25 milligrams of
so-called 1,000 activity. The celluloid disks were coated with gel-
atin containing radium of 10,000 and 25,000 so-called activities.
These latter coatings are said to possess the advantage of yielding
the maximal action of all the rays emitted, since they can be brought
into immediate contact with the skin without any intervening sub-
stance which may absorb the less penetrating rays. As far as could
be ascertained with the apparatus at hand, and through the use of
the electroscope, no loss in activity took place from the coatings
after they had been in contact with the underlying tissues, although
the tissues themselves were proven to have become radio-active.
The coatings were bound by means of plaster strips to the skin cov-
ering the tumors, and the capsules were fitted into small felt pads
possessing circular openings and were also bound in the same manner
"Bashford, Action of Radium on Transplanted Mouse Tumors and Its
Relation to the Spontaneous Arrest of Their Growth, Scientific Report of the
Imperial Cancer Research Fund, 1905, No. 2, Part II, p. 56.
Digitized by
Google
76 Influence of Badium Bromide on Tumor of Bat.
to the skin. The period of radiation ranged from five minutes to
six hours, and there was also considerable variation in frequency
of treatment.
The rat tumors subjected to radiation belonged to the transplan-
tation generations between the ninth and the seventeenth, inclusive,
and corresponded to the stage of simple carcinoma and adeno-car-
cinoma. The method was to select a number of inoculated animals
for radiation, and to retain a similar number of animals in a corre-
sponding condition, as controls. The size of the tumors at the
beginning of the radiation varied from 0.5 to 2 centimeters in
diameter.
It is noteworthy that the effects of the first radiations were very
severe, not on the tumor so much as on the general condition of
the animals. That is to say, all the rats subjected to radiation in
the first few days of the experiment died. This result was attrib-
uted to the fact that the radium had been lying undisturbed in a
leaden box for a period of some months, and there had, therefore,
been an accumulation of rays and possibly the alpha rays especially,
which proved injurious in the manner indicated. A similarly unfor-
tunate result followed the first application of the celluloid coatings,
which had also remained unused for several months. Later on,
after the treatment had been inaugurated, no such deleterious effects
were encountered.
In all but three of the animals submitted to radiation with the
Lieber specimens the nodules continued to grow. Complete dis-
appearance occurred in three instances after radiation with the
1,000,000 specimen. These animals showed on subsequent post-
mortem examination normal organs. At first sight it might appear
that the disappearance of these tumors was referable to the radiation,
but in view of the considerable tendency to spontaneous retrogres-
sion exhibited by this tumor this assumption is not wholly justified.
Indeed, the decision in this case would have to be made, not from
the number of retrogressions occurring, but rather from the number
of tumors which continued to extend. Viewed in this light, the
inevitable decision would be that the radiations had not produced
the disappearance of the tumors.
There is one cause of temporary diminution in the size of the
Digitized by
Google
Maud L. Menten. 77
tumors subjected to radiation that should be mentioned. When
ulceration takes place there occurs frequently an apparent diminu-
tion in the size of the nodules, but this is produced through the loss
of tissue from sloughing and subsequent collapse of the peripheral
thin shell remaining. The ulceration tends to occur earlier in the
radiated than in the non-radiated nodules, probably because of the
injury to the skin produced by the radiation. Now the tumors
which have undergone this early ulceration, far from being retarded
in their growth, continue to extend actively at the borders and ulti-
mately to produce metastases and to cause death.
A series of experiments was made on animals which had received
the tumor inoculations on the two sides of the body. The purpose
of the experiments was to ascertain if the radium when applied to
the tumor on one side would exert any influence on the growth of
the tumor on the opposite side of the body, or if the structure of
the tumors on the two sides would exhibit any marked differences.
The results of the experiments indicated that no special influence
was exerted by the radiation either on the tumor immediately treated
or on the one on the opposite side. The relative rates of growth
showed nothing that the controls did not also show, and the micro-
scopical characters .of the two sets of tumors were also in agreement.
At the conclusion of the first series of experiments, which were
quite negative in result, we secured through the kindness of Dr.
Abbe the use of a preparation of radium much stronger than any
which we had previously employed. This specimen consisted of a
hard rubber capsule as described, containing ten milligrams of
1,800,000 radio-active radium bromide. This constitutes what Dr.
Abbe calls a standard cell. Through its use and his aid by the use
of his photographic method we were able to standardize the various
preparations loaned us by Mr. Lieber. Thus the specimen contain-
ing ten milligrams of 1,000,000 radium was shown to be % of the
standard strength, and the twenty-five milligrams of 10,000, %5 of
of the standard strength. The coatings of 10,000 and 25,000 radio-
activity were respectively about %oo and Hoo of the standard cell.
To obtain the theoretical effects of the standard cell it would there-
fore be necessary to continue the radiations with the weaker
Digitized by
Google
78 Influence of Radium Bromide on Tumor of Eat.
preparations as many times longer as they are weaker than the
standard cell.
Twenty- four rats possessing tumors were subjected to radiation
with Dr. Abbe's standard cell, of which nineteen were radiated for
one hour, two for two hours, one for three hours, one for one hour
and twenty-five minutes, and one for one hour and twelve minutes.
The size of the tumors in this series of rats at the time of the radia-
tion ranged from 0.4 by 0.6 centimeters to i.o by 1.6 centimeters.
The first noticeable effect of the radiations was the loss of hair
over the radiated area, which was accompanied by a severe derma-
titis, followed from ten days to two weeks later by ulceration in
several nodules. The area of necrosis and original ulceration cor-
responded exactly to the dimensions of the overlying exposed sur-
face of the radium. The ulcers, once formed, extended afterwards.
In six of these twenty-four animals the tumors underwent shrink-
age, and in four of these six, in which the tumors were much smaller,
they completely disappeared. To some of the animals which sur-
vived and in which retrogression did not take place, a second radia-
tion of one hour was given a month later, but no influence was
exerted on the progress of the tumors. The results of this experi-
ment must be interpreted in the manner of the last, that is to say,
the retrogressions were not more numerous than often occur spon-
taneously, and the fact that so many of the tumors continued to
extend in spite of the radiations indicates that this tumor when
developing beneath the skin is not subject to marked inhibition of
growth through the radiation.
In order to represent the experiments which were carried out,
and the results obtained, the number of animals in each experiment,
as well as the nature of the experiment, will be stated.
Nine rats were treated with the radium coatings. The radia-
tions were repeated weekly during the life of the animals. In no
case did the tumors disappear, and in every case but one in which
there was survival beyond the first few days following the first
radiation there was increase in size. In one instance a partial
shrinkage of the tumor took place, but complete disappearance never
occurred.
Nineteen rats were submitted to radiation with the capsule of
Digitized by
Google
Maud L. Menten. 79
10,000 radium bromide. The nimiber of exposures varied from
one to twenty-four. In two animals only did the tumors disap-
pear. In all the others they continued to extend. Hence no effect
on the development of the tumor was exerted by these radiations.
Eighteen rats were submitted to radiation with the 1,000,000
specimen of radium bromide. The radiations ranged from one to
ninety in number, and from a few minutes to several hours in dura-
tion. In one instance only was there complete disappearance of the
tumor, and in four instances was there practical arrest of the growth.
Twenty- four rats were submitted to radiation with the 1,800,000
specimen of radium bromide. In four rats the ttimors completely
disappeared and in two others there was arrest of development. In
the others the tumors extended until the death of the animals.
Six rats in which tumors were growing on both sides of the body
were submitted to radiation of the tumor on one side. The number
of radiations varied from two to fifteen. The radium employed
consisted of the 10,000 and 1,000,000 capsules. No appreciable
healing effects were produced.
The almost uniformly negative results obtained with the radium
in the treatment of the rats led to the testing of the radium prepa-
rations on certain mouse timiors. The Jensen tumor, which had
previously been studied in this respect by Bashf ord, was first chosen.
Eight mice, for which a number of control animals were kept, were
submitted to radiation with the several Lieber preparations, for
varying periods of time. In six of these eight the tumors com-
pletely disappeared in from two weeks to two months. The two
tumors which had not completely disappeared were found on micro-
scopical examination to be completely necrotic. No living cells
were found in the tumor nodules proper. The next experiments
with mice were carried out with animals in which the Ehrlich sar-
coma was developing. The Abbe standard cell was employed, and
the radiation continued for one hour. No appreciable effect was
exerted upon this tumor, which continued to grow in all the animals
as it did in the controls. On subjecting these tumors to micro-
scopical examination, no differences in structure as compared with
the unradiated specimens could be discovered.
Digitized by
Google
80 Influence of Radium Bromide on Tumor of Rat.
CONCLUSIONS.
The carcinoma of the rat with which we have dealt, when devel-
oping in the form of nodules beneath the skin and near the surface,
is not subject to inhibition of growth and such injury as tends to
produce retrogression through the influence of radium emanations.
White rats are, on the other hand, highly susceptible to the injurious
effects of certain of the radium emanations, which may produce
death within a short time following the exposure. The local
injurious effects of the emanations, when producing no general dis-
turbance in the condition of the animals, suffice often to cause fall-
ing out of the hair, the development of a severe dermatitis and
ulceration of the skin and the tumors. The extension of the tumors
at the edges of the ulcer is not materially restrained.
The Jensen mouse tumor has been found by us as by others to be
readily subject to the injurious effects of the radium emanations,
which act upon it in such a manner as to lead to its disappearance
without causing appreciable injury to the general health of the mice
in which the tumor is growing. On the other hand, the Ehrlich
spindle-cell sarcoma of mice is not subject to the inhibiting influence
of the radium emanations, but continues to grow without apparent
diminution of energy, in a manner similar to the control animals not
submitted to radiation.
Digitized by
Google
SPONTANEOUS TUMORS OF THE MOUSE.
By J. W. JOBLING, M.D.
(From the Laboratories of the Rockefeller Institute for Medical Research,
New York.)
Plates XVII-XXVIII.
Since Hanau* first showed, with a carcinoma of the rat, the
possibility of successful transplantation of a malignant tumor from
one animal to another of the same species, numerous observations
have been made along similar lines. ^ Morau* showed that the
mouse was subject to transplantation of carcinomatous tumors, and
Loeb* demonstrated that other carcinoma and sarcoma could be
successfully transplanted in rats. Jensen,*^ and later Loeb, carried
out careful histological examinations of fragments of transplantable
mouse tumors, removed at varying periods following inoculation, and
ascertained that the new development of the tissue took place en-
tirely from the transplanted epithelial cells and not from the tissue
of the host. Bashford® has not only confirmed these observations,
but he has shown conclusively that the stroma of the tumors is the
element that is derived from the host. In recent years an immense
activity has developed in connection with the experimental study
of transplantable tumors, and many facts of fundamental biological
importance have been settled in the course of this experimental
investigation. In the following paper no special reference will be
* Hanau, A., Fortschr. d, Med., 1889, vii, 321.
' It is historically of interest to allude to the unsuccessful attempt of Joseph
Leidy to transplant a human mammary carcinoma into the frog, reported by him
to the Academy of Natural Sciences of Philadelphia in 185 1 (Proc, v, p. 201).
Four pieces of the tumor, each half an inch long by one-eighth of an inch
broad and thick, were inserted beneath the integument of the back of a large frog.
•Morau, H., Arch, de mid. exper. et d'anat, path., 1894, vi, 677.
* Loeb, L., Jour, of Med. Research, 1901, vi, 28.
'Jensen, C, Cent, f. Bakt., Orig., 1903, xxxiv, 28, 122.
* Bashf ord, E., Scientific Report of the Imperial Cancer Research Fund, 1905,
No. 2, 24.
81
Digitized by
Google
82 Spontaneous Tumors of the Mouse.
made to the general lines of investigation being followed, but there
will be described, as briefly as possible, the histological structure of
and the results obtained from the transplantations into mice of a
considerable number of tumors developing spontaneously in that
class of animals. During the past three years there have come into
our hands twenty-six mice showing spontaneous tumors. These
twenty-six animals showed, according to conclusions based upon
my study, forty-one primary tumors. In one mouse there was
present what appeared to be either a general hyperplasia of lym-
phoid tissue, or else many metastases from a large superficial lym-
pho-sarcoma, but, of course, this animal was taken as representing
one tumor formation. However, ten of the mice, or 38.4 per cent,
of the entire number, showed more than one tumor each. Three
of these showed three tumors each, and seven, two tumors each.
If we compare our observations with others reported in the liter-
ature, we shall find that they agree generally with previous obser-
vations, although differing considerably in detail. Thus Apolant^
reported that among 221 mice showing spontaneous tumors he
found 276 primary tumors; that is, 12 per cent, of the mice men-
tioned possessed more than one tumor. Tyzzer® has also reported
multiple primary tumors in mice, and Murray^ mentions that of
119 mice with spontaneous tumors 142 tumors in all were detected.
In other words, in Murray's series 15 per cent, of the animals pos-
sessed more than one tumor. Reporting further, he states that
fourteen mice showed two tumors each; three, three tumors each;
and in one animal the entire inguinal mamma had been transformed
into a group of nodules. In six of the animals of our series*®
there were present large subcutaneous tumors, and in six, what were
taken to be primary tumors of the lung. One mouse showed a sub-
cutaneous sarcoma, a cyst-adenoma of the ovary, and an adenoma-
tous tumor of the lung; another mouse showed a superficial
^Apolant, A., Arh, a. d. k, Inst. /. exper. Therapie, Frankfurt, 1906,
No. I, II.
• Tyzzer, E., Fourth Report of the Caroline Bretver Croft Fund Cancer Com-
mission, 1907, 27.
•Murray, Scientific Report of the Imperial Cancer Research Fund, 1908,
No. 3, 69.
*• Jobling, Proc. of the Soc. for Ex per, Biol, and Med., 1908, vi, 10.
Digitized by
Google
/. TT. Johling. 83
sarcoma together with cyst-adenomata of both ovaries; and, finally,
two mice each presented two subcutaneous tumors of different
t3rpes.^^
TRANSPLANTATION.
The voluminous literature which has already accumulated on the
transplantation of mouse tumors contains many statements of suc-
cessful and unsuccessful transplantation, but there are available
relatively a small series of observations which enable us to deter-
mine the percentage of transplantable tumors. Ehrlich^^ reports
upon 230 tumors, of which 94 were inoculated into other mice, and
II per cent, of these proved to be transplantable. Tyzzer^^ re-
ports four carcinomata of mice, two of which were successfully
transplanted. Bashf ord^* reports that 1 5 out of 32 sporadic tumors
transplanted gave negative results. Of our series of 41 tumors, 26
were inoculated into other mice, and of these 20, or yy per cent.,
developed. From these timiors 1,128 mice were inoculated, of
which 855 survived more than two weeks. Of these, 104, or 12.1
per cent, developed tumors. Of these 104 grafts which began to
grow, 36, or 34 per cent., later underwent retrogression. The num-
ber of tumors developing in the first generations varied from 2 to
46 per cent, and the period of incubation of the first tumors ranged
from 30 to 180 days. In Ehrlich's series 1,504 mice were inoculated
with material from 94 primary tumors, 41 of which, or 2.7 per cent.,
developed tumors. The percentage of tumors developing in the
first generation of our different implantations ranged from 2 to 50
per cent. Bashford^*^ transplanted 32 spontaneous tumors of the
" Tyzzer in a later article (Boston Med. and Surg, Jour., 1909, dxi, 103) says :
" DiflFerent types of tumors frequently occur in a single animal. Of the 49
animals of this series 11 presented primary tumors of two types, and i animal,
primary tumors of four types. In the last there were a hypernephroma, a lympho-
sarcoma, a papillary cyst-adenoma of the lung, and an adeno-carcinoma of the
ovary."
" Ehrlich, P., Arb. a. d. k. Inst f. exper, Therapie, 1906, No. i, 77.
*■ Tyzzer, E., Fourth Report of the Caroline Brewer Croft Fund Cancer Com-
mission, 1907, 27.
"Bashford, Scientific Report of the Imperial Cancer Research Fund, 1908,
No. 3, 313.
"Bash ford. Scientific Report of the Imperial Cancer Research Fund, 1905,
No. 2, Part II, 19 and 30.
Digitized by
Google
84 Spontaneous Tumors of the Mouse.
mamma into 2,278 mice, which survived a sufficient time to permit
of a final estimation, and 72 tumors developed. That is, one inoc-
ulation in 3 1. 1 mice was successful, yielding 3.2 per cent, of
successes.
LOCALITY.
In describing the location of the spontaneous tumors of 221 mice,
Apolant^® states that 25 per cent, were situated on the side of the
abdomen, 16 per cent, in the region of the vulva, and 15 per cent,
between the sternum and the submaxillary glands. Tyzzer^^ re-
ports 20 cases, in which 5, including one lympho-sarcoma, were
located in the subcutaneous tissues of the abdomen, 12 were primary
lung tumors, one was a lympho-sarcoma of the thorax, and two
were adenomata of the kidneys.** Murray*® does not describe ac-
curately the frequency with which the tumors were found in differ-
ent parts of the body, but judging from the diagram prepared by
him and showing the positions of 142 primary tumors, the great
majority developed on the side of the thorax and in the inguinal
regions.
In our series^^ of 41 primary tumors, 9 appeared in the lungs, 3
in the ovaries, and 29 in the subcutaneous tissues. Of the last, 4
were located in the neck, 9 on the side of the thorax, 14 in the in-
guinal region, and 2 in the region of the vulva. This series does
not include the mouse in which there were multiple growths com-
posed of lymphoid tissue.
"Apolant, A., Arb. a. d. k. Inst, f, exper. Tkerapie, 1906, No. i, 11.
" Tyzzer, E., Fourth Report of the Caroline Brewer Croft Fund Cancer Com-
mission, 1907, 27.
"In a later paper (op, cit.) Tyzzer says: "I have now studied a series of 62
primary tumors in mice, and of these, 37, or about 60 per cent., originated in the
lung. The next most frequent type of tumor in my series is the lympho-sarcoma,
of which there were 10. Only 8 of the 62 tumors were situated externally and
6 of these were epithelial tumors. There occurred 4 tumors of the kidney, of
which 2 were undoubtedly hypernephroma. In 2 mice ovarian tumors occurred
and in i mouse, a sarcoma."
" Murray, Scientific Report of the Imperial Cancer Research Fund, 1908, No.
3, 69.
* Jobling, Proc. of the Soc, for Exper. Biol and Med., 1908, vi, 10.
Digitized by
Google
J. W, Jobling. 86
METASTASIS.
Hanau*^ in his original article reported the existence of metatases
in the lymphatic glands in the neighborhoood of the cancer of the
rat. Similar metastases, either appearing with the primary tumor,
or as a result of transplantations, have been described by Loeb,^'
Borrel,^* Haaland,^^ Apolant,^ Flexner and Jobling,^® Michaelis,^^
and Tyzzer.^® With few exceptions the metastases have been con-
fined to the lungs. Apolant reports that of his series of 221 mice
suffering from spontaneous tumors, metastases in the lungs were
found five times, or in 2.2 per cent, of the animals. Murray**
reports that of 68 mice with spontaneous tumors, 27 presented lung
metastases, and 3, lymphatic gland metastases. In 26 mice of our
series lung metastases were found in 5, or in about 20 per cent.
of the animals, and no metastases were discovered in any other
organ of the body.
MAMMARY TUMOR.
Of the 41 tumors, 29 developed in positions of the body corre-
sponding with the distribution of the mammary glands, and these,
with the exception of 5, all presented well defined adenomatous
structure. We have concluded in conformity with the present view
of different workers that these tumors originated from the mam-
mary gland. Attempts have been made to classify these tumors
according to types. Michaelis*^ distinguishes three different types,
and Haaland,*^ four; but probably the most satisfactory classifica-
tion at present is that offered by Apolant.^*
"Hanau, A., Fortschr. d. Med., 1889, vii, 321.
"Loeb, L., American Jour, of the Med. Sciences, 1903, cxxv, 243.
"Borrel, Ann. de VInst. Pasteur, 1903, xvii, 81.
•* Haaland, Ann. de VInst. Pasteur; 1905, xix, 165.
"Apolant, Arh. a. d. k. Inst, f, exper, Therapie, 1906, No. i, 11.
"Flexner and Jobling, Jour, of the American Med. Assn., 1907, xlviii, 420.
* Michaelis, Zeit. f. Krebsforsch., 1907, v, 189.
" Tyzzer, E., Fourth Report of the Caroline Brewer Croft Fund Cancer Com-
mission, 1907, 27.
" Murray, Scientific Report of the Imperial Cancer Research Fund, 1908, No.
3, 69.
•• Michaelis, Zeit, f. Krebsforsch., 1906, iv, i.
"Haaland, M., Ann, de VInst, Pasteur, 1905, xix, 165.
"Apolant, H., Arb, a, d, k. Inst, f. exper, Therapie, 1906, No. i, 11.
Digitized by
Google
86 Spontaneous Tumors of the Mouse.
He begins by referring to the difficulty of classifying these mouse
tumors, and speaks of the similarity of their histological structure
to thyroid tumors of man. He distinguishes two main groups,
adenoma and carcinoma, and further divides the adenoma into
simple adenoma, cyst-adenoma, hemorrhagic cyst-adenoma and
papillary cyst-adenoma. The carcinoma he divides further into
simple alveolar carcinoma, hemorrhagic cyst-carcinoma, and papil-
lary carcinoma. We have followed Apolant's classification in de-
scribing the twenty-five tumors of our series, composed of epithe-
lial cells and stroma, and arising in the mammary region. These
twenty-five tumors were composed of ten hemorrhagic cyst-ade-
nomata, twelve adeno-carcinomata, two cyst-adenomata, and one
alveolar carcinoma. Before proceeding to a somewhat more de-
tailed description, it should be stated that these tumors are not strictly
homogeneous, but that almost every one shows heterogeneous areas
of a type differing more or less from the one under which it was
classified, so the most that can be said is that the dominant structure
has been chosen for the purpose of classification.
ADENO-CARCINOMA.
By far the greater number of the tumors of mice thus far re-
ported belong to this class. It is interesting to recall that Hanau's^*
original study was of a squamous cell carcinoma of the vulva of the
rat, but the tumors of mice and rats described since that time have
been with few exceptions adeno-carcinomata of the mammary
gland. Of the twelve spontaneous tumors of this class coming
into my hands, ten were used for transplantation, of which nine
were successfully transplanted into other mice. Of 464 mice inoc-
ulated with these tumors, 407 survived two weeks or longer, of
which 57, or 14 per cent., developed tumors. The average period
of incubation for this series was 63 days. Among the ten animals
whose tumors were transplanted, four showed lung metastases and
four also exhibited primary tumors of the lung.
These tumors were, with one exception, in which there was ulcer-
ation of the skin, freely movable and apparently encapsulated.
The tumors in two of the series showed areas of keratinization.
•" Hanau, Fortschr. d. Med., 1889, vii, 321.
Digitized by
Google
J. W. JobUng. 87
The keratin transformation of the epithelial cells of these tumors
has long been known. It was present in Hanau's original carci-
noma, and has been described by Borrel and Haaland,^* Erdheim,**^
Haaland,^® Bashford^'' and Lewin.'® In Lewin's case, as in
Hanau's, the condition was present in tumors of the rat, and in all
the others, in tumors of mice. In Lewin's case the keratinization
did not appear until the third generation, when it became diffuse.
In only one of the two cases observed by us did the transplanted
fragments develop, and it is interesting to record that the keratiniza-
tion did not appear in the subsequent tumors.
Generally speaking, these tumors are solid in structure and gray-
ish-white in color. They are composed of microscopical lobules
filled with closely packed, regular acini, which often contain
a material of homogeneous structure staining deeply in eosin.
That these tumors are to a certain extent invasive is shown by the
presence of striated muscle fibers in the main tumor. The stroma
separating the lobules tends to be dense and thick, although cellular
in places, while that separating the acini is delicate and sometimes
difficult to make out. The acini are usually lined with a single
layer of cells, cubical in shape, and possessing nuclei rich in chro-
matin. All the tumors of this class show in addition to the acinous
arrangement solid nests of cells. These cells are no longer cubical,
but irregularly shaped, and possess a large vesicular nucleus, which
is less rich in chromatin. These cells also invade the acini, which
they do not quite fill. The two kinds of cells are readily differen-
tiated by their staining characters.
DESCRIPTION OF THE INDIVIDUAL TUMORS.
I. The first of this series was in an old, white, female mouse
(Original iv). The tumor occupied the left inguinal region,
measured three and one-half centimeters in diameter and had in-
vaded the lumbar and abdominal muscles. The skin was involved
••Borrel and Haaland, Cotnpt. rend. Soc, de biol, 1905, Iviii, 14; Haaland,
Ann. de VInst. Pasteur, 1905, xix, 165.
" Erdheim, Zeit. f, Krebsforsch., 1906, iv, 33.
" Haaland, Norsk Mag, for Lcegesvidensk., 1907, v, 105.
''Bashford, Murray and Haaland, Berliner klin, Woch., 1907, xliv, 1194.
" Lcwin, Berliner klin, Woch., 1907, xliv, 1602.
Digitized by
Google
88 Spontaneous Tumors of the Mouse.
and tilcerated. On section there were small cysts and hemorrhagic
foci. The right lung presented a nodule the size of a pin head.
The structure was characteristic, being chiefly adenomatous (Plate
XVII, Fig. I ), but showing some solid nests of cells and numerous
mitotic figures. As was to be expected, small hemorrhages were
also present. The lung nodule consisted of a metastasis composed
mainly of a solid growth of epithelial cells, with, occasionally, the ap-
pearance of acini. It was situated just beneath the pleura and
showed numerous mitotic figures. The stroma was delicate.
From the original tumor 50 mice were inoculated, of which 3, or
6 per cent., developed timiors. Of these 3, a growth was first noted
in two on the nintieth and in one on the one hundred and fiftieth day
after implantation. The transplantation tumors were made up of
solid growth of tumor cells (Plate XVII, Fig. 2), between which the
stroma was slight and cellular, and sometimes edematous. There
were numerous mitotic figures, many vessels, and frequent hemor-
rhages. The necrotic areas were frequent.
In the succeeding generations the tumor preserved its histological
characters, but in the sixth generation the necrosis and hemorrhages
became more numerous, and acini, similar to those in the original
tumor, became common (Plate XVIII, Fig. 3). The successful
transplantations of this tumor never exceeded 40 per cent., and the
incubation period averaged ultimately about fifteen days.
2. An old, white, female mouse (Original vi) showed a tumor
measuring one and one-half centimeters situated in the left in-
guinal region. The skin over the tumor was tense and purple, but
free. The tumor was movable. On section there were a few
hemorrhagic areas in an otherwise grayish-white tissue. The right
lung showed a tumor nodule.
The inguinal tumor was of glandular type, consisting of tubules,
some of which were dilated to form cysts, and there were associated
solid areas of cells (Plate XVIII, Fig. 4). The hemorrhagic cysts
were apparently of secondary origin, and caused by hemorrhage into
necrotic foci. The lungs showed several small metastases confined
about the blood vessels and composed of solid nests of epithelial
cells. In one instance there was observed tubular formation within
a small blood vessel. In one place a growth was observed in an
Digitized by
Google
J. W. Jobling. 89
artery, from which it extended into branches, and in another place
several large arteries were occluded by the growth. There were
numerous mitotic figures present.
59 mice were inoculated, of which 3, or 5 per cent., developed
tumors. The average period of incubation was 120 days. Under
the microscope the secondary tumors resembled one another, but
differed greatly from the original tumor. The adenomatous struc-
ture so striking in the original had been succeeded by solid growths
of cells in the secondary tumors (Plate XIX, Fig. 5). Mitotic fig-
ures were numerous, and hemorrhages were common in the stroma
and between the tumor cells. With material from the first trans-
plantation generation, 143 mice were inoculated, of which 8, or 5.6
per cent., developed tumors. This tumor was carried to the 7th
transplantation generation, and the number of successes never ex-
ceeded 50 per cent. The microscopical structure of the transplan-
tation generations remained essentially fixed, and consisted chiefly
of solid nests of cells with here and there acini and a small amount
of stroma. The only striking variation was a somewhat greater
development of the cellular and edematous stroma. Hemorrhages
appeared not infrequently.
In the second generation one of the tumors presented an appear-
ance somewhat suggestive of a mixture of sarcoma and carcinoma
(Plate XIX, Fig. 6). The epithelial cells formed lobules, the central
portions being usually necrotic. The lobules were separated
from one another by thick bands of very cellular stroma.
Large blood vessels were present, the walls of which, in many in-
stances, were composed only of a layer of endothelial cells. It was
at first thought that we were dealing here with a transition of the
carcinoma into a sarcoma, but the daughter tumors which developed
gave no further evidences of this structure.
3. An old, white, female mouse (Original vii) showed a large
semi-fluctuating tumor in the right inguinal region. The skin was
tense and purple in color, but free. On section the tumor was
composed simply of two large cysts filled with a dark brownish
fluid. The tissue at the margins was of a grayish color and friable.
Under the microscope the preserved tissue showed a structure simi-
lar to No. 2.
Digitized by
Google
90 Spontaneous Tumors of the Mouse.
From this tumor i8 mice were inoculated, of which 3, or 16.6
per cent., developed tumors. The incubation period for two of the
three was 45 days, and for the other it was 70 days. Under the
microscope these tiunors were seen to be composed of closely packed
acini, and small hemorrhages were observed. For the second trans-
plantation generation, 61 mice were employed, and 9, or 14.7 per
cent., developed tumors. The shortest incubation period was 30
days, and the longest was 1 50. For the third generation, 56 mice
were inoculated and 4, or 7.1 per cent., developed tumors. Of
these 4, 3 subsequently were completely absorbed. The histologi-
cal structure of the transplantation generations was similar to that
of the original tumor. The acinous arrangement recurred, but
cysts were also present. Mitotic figures were numerous.
4. A very old, white, female mouse (Original xii) presented a
tumor measuring one and one-half centimeters in the right inguinal
region. The tumor was freely movable. In consistence it was
quite firm. The left lung presented at the lower margin a grayish-
white wedge-shaped tumor measuring three millimeters in diameter.
In the right axilla was a nodule measuring 0.5 centimeter. The
original ttmior showed under the microscope a somewhat compound
structure. While the main portion was composed of solid growths
of epithelial cells, there were some acini and small dilated cysts.
Where the glandular type occurred, the stroma was fibrous, while
in the more solid portions it was delicate. There were some hemor-
rhages, and mitotic figures were numerous. The lung nodule
proved to be a primary cyst-adenoma and will be described else-
where.
The nodule in the right axilla proved to be composed of the
closely packed tubules separated from one another by delicate
stroma. The nuclei of the small cells were dense and mitotic
figures were not found. No remnants of lymphoid tissue were
discovered, and it is probable that this represented an independent
tumor formation.
50 mice, 30 white and 20 colored, were inoculated, of which 33
survived two weeks or longer, and one developed tumor, which,
however, subsequently underwent absorption. It was detected for
the first time 30 days after the inoculation.
Digitized by
Google
/. W. JobUng. 91
5. An old, white, female mouse (Original xiii) showed a tumor
measuring one and one-half centimeters in the left inguinal region,
which was freely movable. The consistence was firm and the cut
surface grayish-white in color. Under the microscope this tumor
was almost of pure glandular type (Plate XX, Fig. 7). Micro-
scopical examination of the lungs showed a primary cyst-adenoma
which will be described elsewhere.
In the first generation, 80 mice were inoculated and 67 survived.
8 of these, or 11.9 per cent., developed tumors, two of which sub-
sequently underwent absorption. The average incubation period
was 70 days. In the second generation, 20 mice were inoculated,
and 6, or 33 per cent., developed tumors. The incubation period was
30 days. The structure of the transplantation tumors was similar
to the original. In the third generation, 20 mice were inoculated,
of which 6, or 33 per cent., developed tumors, and the incubation
period was 25 days. In this generation the histological type had
altered, so that some of the tumors corresponded to the original, and
others were composed of solid lobules with necrotic centers and
heavy stroma (Plate XX, Fig. 8). In both, mitotic figures were
numerous.
6. A Japanese waltzing mouse (Original xv) presented a tumor
situated in the front of the neck, almost in the median line, freely
movable, and measuring one centimeter in diameter. Under the
microscope this tumor was made up of numerous acini and a few
cysts. The cells forming the more solid portions had through
pressure become flattened and spindle-shaped. The stroma was
slight, and mitotic figures were nimierous. Two small metastases
containing acini, but less uniform in structure than the original
tumor, were present in the lungs.
40 mice, consisting of 30 white and 10 waltzing mice, were inocu-
lated. None of the white mice developed tumors. Seven of the
dancing mice survived longer than two weeks, of which three, or
42.8 per cent., developed tumors. These three died while the
tumors were still small, and no further transplantations were at-
tempted. The transplantation tumors resembled histologically the
original.
7. A white, female mouse (Original xx), showed, in addition to
Digitized by
Google
92 Spontaneous Tumors of the Mouse.
two adenocarcinomata of the type we are describing, other primary
tumors, namely, two hemorrhagic cyst-adenomata of the neck and
inguinal region, and a primary cyst-adenoma of the lung. One of
the adeno-carcinomata, situated on the lower part of the abdominal
wall, measured two centimeters, and the other, which was situated
in the left inguinal region, measured seven millimeters. These
tumors presented a distinctly glandular character, being composed
of acini similar to the others of the series. Muscle fibres had in
some places been included in the growing tumors.
40 mice were inoculated, of which 28 survived, and 12, or 42.8
per cent., developed tumors. The incubation period ranged from
30 to 60 days. The tumors presented microscopically two types:
first, similar to the original ; and second, solid cell nests devoid of
acini. 40 mice were inoculated in the second generation, and
yielded 16, or 40 per cent, of tumors, the average incubation period
being 30 days.
8. An adult, light brown, female mouse (Original xxiii) showed
a tumor measuring one by one and one-half centimeters, in the right
inguinal region, surrounded by subcutaneous fat. It was solid in
consistence. In microscopical structure the tumor was of the glan-
dular type. It contained small cysts, and the stroma was fairly well
developed. Mitotic figures were numerous.
26 white (?) mice were inoculated, of which 2, or 7.6 per cent.,
developed tumors, one appearing at the end of 60, the other at the
end of 120 days. Except for more numerous hemorrhages, they
resembled the original. In the second generation, 20 mice were
inoculated, of which 3, or 15 per cent., developed tumors. The
structure had now altered : there were fewer acini, a general solid
growth of cells, and numerous hemorrhages. Mitotic figures were
numerous.
9. A white, female mouse, apparently very old (Original xxiv)
showed a tumor measuring 1.3 centimeters on the right side of the
thorax, just below the axilla. The structure was solid and lobu-
lated and there was involvement and ulceration of the skin. Under
the microscope the growth was more solid in type than the other
tumors of this series, but a certain amount of acinous formation
could be made out. Keratinized foci (Plate XXI, Fig. 9) were num-
Digitized by
Google
/. W. Jobling. 93
erous, the keratinized cells sometimes appearing at the margin of a
necrotic area, sometimes surrounded by necrotic material, and some-
times occurring in the preserved parts of the tumor.
50 mice were inoculated with this tumor, but no secondary tumors
developed.
ID. A yellow, female mouse (Original xxv) presented a tumor
measuring one-half centimeter in diameter, situated on the right
side of the neck and freely movable. The tumor was friable, and
showed minute hemorrhages. There were numerous nodules in the
lungs. The microscopical structure was acinous and similar to
others in this series, and there were numerous small cysts present
containing a serous exudate and sometimes blood. Within some
of the cysts there were keratinized areas, showing typical " pearls "
(Plate XXI, Fig. 10). Some solid areas of cells also occurred. The
lung metastases were numerous and were confined to the blood
vessels, which, in many instances, were occluded by them. In
other cases the lung tissue was replaced. When the growth oc-
curred in the blood vessels it was usually solid, and mitotic figures
were numerous. When the growth in the vessel did not fill it, the
outer surface was sometimes covered by endothelium, and in one
instance a second independent growth of tumor was observed to
have taken place between the old mass, which was covered with
endothelium, and the vessel wall. The central portions of some
of the large intravascular growths had become necrotic and formed
cavities. The extra-vascular growths showed acini and some solid
nests and numerous mitotic figures (Plate XXII, Fig. 11).
There was a larger, wedge-shaped nodule in the lung, which
proved to be a primary cyst-adenoma, and which will be described
elsewhere (Plate XXII, Fig. 12).
Of 50 mice inoculated, 23, or 46 per cent., developed tumors.
Of 33 mice inoculated in the second generation, 18, or 54.5 per
cent., developed tumors. In structure the secondary tumors resem-
bled the original, except that keratinization was not observed.
HEMORRHAGIC CYST-ADENOMA.
A somewhat special interest is attached to this class of tumors
because Ehrlich*® employed them in an extensive series of experi-
•• Ehrlich, Arb. a. d. k. Inst. f. exper. Therapie, 1906, No. I, 77.
Digitized by
Google
94 Spontaneous Tumors of the Mouse.
ments to induce immunity. Having discovered that they were
transplanted with greater difficulty than other mouse tumors, and
that animals which had successfully resisted their implantation
presented an increased resistance to the inoculation of other tumors
readily transplantable to normal mice, he employed them for pro-
ducing this refractory or immune state. On the other hand, some
investigators have succeeded in transplanting this type of hemor-
rhagic tumor. Hertwig and Poll*^ report a successful instance,
and Tyzzer^^ reports a successful instance, so far as the first genera-
tion is concerned, while Gierke*^ reported that of 47 such tumors, 38
were transplanted successfully by him. He gives the details of his
experiments with 35 separate tumors. Of these 35, 29 were trans-
plantable, and of the 29, 15 became extinct as follows : 8 in the first
generation, 5 in the second, and 2 in the third. In only one in-
stance was he successful in carrying a tumor of this class into the
fourth transplantation generation. Gierke states that of 2,851 mice
which survived for a sufficient time, 187, or 6.5 per cent., developed
tumors. The highest original result ever attained was 33 per cent.,
and the highest subsequent one was 70 per cent, of the tumor frag-
ments implanted.
The tumors of this class can frequently be detected from their
gross appearances, since the skin covering them is usually tense
and purplish in color, and the section of the tumor shows large
cysts containing bloody contents and little original tumor tissue.
They are as a rule incapsulated. Under the microscope the more
solid portions are composed of regular acini, but everywhere there
are cysts of variable size, produced through dilatation of the acini
and through necrosis of the tumor. The cells which line the acini
are usually cubical and rich in chromatin, while those lining the cysts
are usually flattened. The stroma is usually delicate in character,
but may be thicker and denser, or edematous. Large hemorrhagic
foci are scattered through the sections, as well as large thin-walled
vessels. The walls of the vessels are so thin at times that they
^ Hertwig and Poll, Ahhand. d. k. preuss. Akad. d. Wissensch, Berlin, 1907, i, i.
*^Tyzzer, E., Fourth Report of the Caroline Brewer Croft Fund Cancer
Commission, 1907, 27.
*• Gierke, E., Scientific Report of the Imperial Cancer Research Fund, 1908,
No. 3, 115.
Digitized by
Google
J. W. JoUing. 95
are made out with great difficulty, and thus it becomes uncertain
sometimes whether the blood is contained within the vessels, or
within a cyst. The mitotic figures are usually infrequent.
In our series there were ten tumors belonging to this class, from
which 365 mice were inoculated. However, as 61 died within the
first two weeks, only 304 survived long enough to permit the devel-
opment of tumors. Of these, 21, or 6.9 per cent., developed tumors
and later seven of the tumors became absorbed. Of these ten
tumors, seven, or 70 per cent., gave transplantation tumors in the
first generation. Of these seven, two failed to develop tumors in
the second generation, and one in the third, while the remaining
four were transplanted successfully through successive generations.
In one of the ten mice a lung metastasis was present, although in
three others primary lung tumors occurred.
1. The tumor measured one and one-half centimeters in diameter,
and appeared in the right inguinal region of a female mouse (Orig-
inal i) kindly presented to us by Dr. F. C. Wood of New York.
The skin covering the tumor was tense, partially free of hair, and
purplish in color. The tumor was fluctuating on palpation, and
on incision, a large amount of dark fluid escaped from a smooth-
walled cyst on the inner wall of which papilliferous masses were
attached.
The solid portion of the tumor was suspended in normal salt
solution, after being cut into minute fragments, and 15 mice were
inoculated under the skin. None developed tumors. Under the
microscope the tumor was found to be composed mainly of cysts
in the tissue, between which there were regular acini. Hemor-
rhages of different durations had taken place into the cysts. A
small number of mitotic figures were seen, and occasionally solid
masses of cells.
2. A brown, female mouse, 15 months old (Original ii), showed
a tumor on the left side of the body, just below the axilla, which
measured 1.75 centimeters. The tumor was grayish-white in color,
lobulated and presented on section many small hemorrhagic cystic
areas. On the opposite side of the body in the same general posi-
tion a nodule 4 by 6 millimeters occurred. It was loosely attached
Digitized by
Google
96 Spontaneous Tumors of the Mouse.
to the skin, and is described under the cyst-adenomata. Under the
microscope the larger tumor was composed of closely packed tubular
acini, lined by a single layer of epithelial cells. Sometimes the
epithelium was arranged in several layers, or entirely filled the
spaces. Hemorrhages were scattered through the tissue, being con-
tained within cyst-like spaces. The blood vessels presented very
thin walls and appeared often as sinusoids. One of the mammary
glands adjacent to the nodule showed much dilated tubules, filled
with a glandular eosin-staining detritus.
41 mice, of which 19 were brown and 22 white, were inoculated
with fragments which had not come into contact with saline solu-
tion, and none of the 19 brown mice developed tumors, while 21
of the 22 white mice developed them. One of these was first
noticed on the isoth and the other on the i8oth day after inocula-
tion. From the former, 20 white mice were inoculated, one of
which died in a few days, and of the remaining 19, 3, or 15.7 per
cent., developed tumors, one of which subsequently underwent
absorption. The incubation period in this series ranged from 60 to
90 days.
Microscopical examination showed the transplantation tumors
to be more solid in character than the original, to contain hemor-
rhages and to show quite numerous mitotic figures. Some acini
were present, and the epithelial cells in the two kinds of structures .
resembled the original. Microscopical examination of the tumors
of the second transplantation generation showed marked changes
as compared with the original. There were still present numerous
acini and cysts, but the latter were produced by a disintegfration of
necrotic cells. The tumors were also more lobulated, and there
were ingrowths from the sides of the lobules of a somewhat variable
character. However, the larger part of the tumor consisted of a
solid growth of cells. The stroma was cellular and well developed
in places, but the blood vessels were still large and thin-walled.
Hemorrhages were not infrequent, but not so common as in the
original tumor. Mitotic figures were not infrequent.
In the third generation, forty mice were inoculated, yielding one
slowly growing tumor. In the fourth generation, twenty were
inoculated yielding seven tumors, or 3S^er cent. Subsequently, six
Digitized by
Google
J. W. Jobling. 97
of these seven were absorbed. Further transplantations were not
attempted.
3. White, female, apparently old mouse (Original iii) showed a
tumor on the left side, just below the axilla and extending in front
almost to the median line. The skin was tense and purplish in
color. On removal the tumor was dark brown, and on section
exuded a dark fluid. It measured three by one and one-half centi-
meters. Under the microscope it proved to be composed of lobules
consisting of closely packed acini. The centres of the lobules usu-
ally were cystic, and the cysts contained an eosin-staining, homo-
geneous substance and blood. Some solid nests of cells existed,
and mitotic figures were numerous. Muscle fibres were contained
in the upper pole of the tumor.
59 mice were inoculated, of which 4, or 6.7 per cent., developed
tumors. The first growth was noticed on the 120th day, and the
later growths on the 150th and i8oth days. All but small parts of
three of the tumors were excised, and used to inoculate 34 mice.
The fragments of tumor left behind underwent subsequent absorp-
tion. Of the 34 mice of this series, one developed a tumor at the
end of 60 days. As the tumor was growing very slowly, an incision
was made in it with the idea that it might be stimulated to more
active growth, but after the operation there was complete absorp-
tion. The microscopical examination of the first transplanted
tumors showed cystic growths in which the cysts were filled with
blood (Plate XXIII, Fig. 13). The general structure was acinous.
The microscopical examination of a fragment removed from the
second generation transplantation showed a papillomatous growth
with cysts and some hemorrhages (Plate XXIII, Fig. 14).
4. White, female, evidently old mouse (Original viii). The
tumor had been partially eaten away by another mouse in the same
cage. It was on the left side and partially surrounded the vagina.
In color it was dark brown, and contained several cysts in which
there was a dark brown fluid. Microscopical examination showed
the tumor to contain a large amount of dense stroma, and to be
divided into small lobules, which themselves were made up of acini.
Hemorrhages were common in and between the latter.
30 mice were inoculated with fragments of this tumor, but no
secondary tumors developed.
Digitized by
Google
98 Spontaneous Tumors of the Mouse.
5. White, female mouse (Original ix) showed a tumor measur-
ing two centimeters on the right side just below the axilla. The
skin was tense and purplish in color. The external portion of the
tumor was composed of a large cyst and there w^ere smaller cysts
in other portions. Under the microscope the tumor was found to
be composed almost entirely of cysts which in places were being
replaced by granulation tissue. The sections removed from the
deeper portions alone showed traces of the original tumor. Many
hemorrhages were present, and the blood vessels were numerous.
40 white mice were inoculated from this tumor, of which 25 died
in less than two weeks. Of the 15 surviving, one, or 6.6 per cent.,
developed a tumor, and the incubation period was 60 days. This
mouse died during the hot summer months, and decomposition set
in so rapidly that no further attempts were made to transplant it
to other mice. Microscopical examination showed the transplanted
tumor to be like the original.
6. White, female and apparently old mouse (Original x) showed
a tumor measuring two centimeters in the left inguinal region.
The skin over it was tense and purplish in color. On section, the
tumor was composed almost entirely of cysts containing a dark
fluid (Plate XXIV, Fig. 15). The cyst walls were thin, and there
was little of the solid tumor tissue left except at the margins. A
small nodule was present just beneath the pleura in the left lung,
which proved to be a primary cyst-adenoma. In microscopical
appearances this tumor was similar to others of this class, and cysts
and hemorrhages were everywhere present. There were numerous
mitotic figures.
30 mice were inoculated, of which 9 died within a few days.
Of the 21 surviving, 2, or 9.5 per cent., developed tumors. The
incubation period was 60 days. One of the two was ultimately
absorbed, and the remaining mouse with tumor died before it be-
came large enough to be used for transplantation.
7. A brown, female and apparently old mouse (Original xi)
showed a tumor at the right axillary region, measuring two by one
and one-half centimeters. It was composed of a large cyst exter-
nally, and was filled with a dark, thick brownish fluid. The deeper
portion was grayish-white in color and firm in consistence. There
Digitized by
Google
J. W. Jobling. 99
were numerous lung metastases. Excepting for the thickness and
density of the stroma, this tumor was very similar to others previ-
ously described. In some places there was a tendency towards
development of papillary outgrowths in the walls of the lobules.
The lobules for the most part were filled with closely packed acini,
and the stroma contained many large phagocytic cells filled with
blood pigment. Mitotic figures were not numerous. The lungs
were thickly studded with metastases, confined chiefly to the blood
vessels. Few of the larger vessels had escaped invasion. The
growths in the lung resembled the original, even to the formation
of cysts and the presence of hemorrhages. This animal died dur-
ing one night and w^hen it was examined the next morning there
was already beginning decomposition. 40 mice, of which 10 were
white and 30 light brown, were inoculated, and no tumors de-
veloped.
8. An adult, white, female mouse (Original xvi) presented a
tumor in the right inguinal region, measuring two and one-half
centimeters. The skin was attached and ulcerated. The tumor
was gray and contained numerous hemorrhages. A single small
cyst was observed. The left lung contained a small primary tumor.
Under the microscope the superficial tumor was distinctly hemor-
rhagic in type, but less homogeneous in structure than the others
of this class. The arrangement of the epithelium was highly
varied. Sometimes it was glandular and formed large cysts; at
other times there were lobules filled with a solid g^rowth. Necrosis
was common, and hemorrhages frequent, and the blood vessels
were large and possessed thin walls. In some places the acini had
become greatly dilated, and into their lumina larger cells with vesic-
ular nuclei had penetrated.
From this tumor 40 white mice were inoculated, of which 12 died
in a few days. Of the 28 surviving, 6, or 21 per cent., developed
tumors ; the incubation period was 20 days. Three of these tumors
subsequently underwent absorption. The microscopical structure
of the secondary tumors agreed with the original.
Of 37 mice inoculated from the first transplantation generation,
I3» or 35 per cent., developed tumors, of which 8 subsequently
underwent absorption. The incubation period was about 25 days.
Digitized by
Google
100 Spontaneous Tumors of the Mouse.
The glandular or acinous structure present in the original tumor
could be made out hardly at all in the tumors of this generation.
The stroma had become well developed, and subdivided the tumor
into small lobules, which were filled with cells showing no definite
arrangement. A striking feature was the strong development of
the stroma. Necrosis of the central portion of the lobule was
common.
20 mice were inoculated for the third generation, of which 5, or
25 per cent., developed tumors. Three of these subsequently under-
went absorption. The incubation period was about 25 days. The
structure remained about the same as the previous generation. 28
mice were inoculated in the fourth generation, of which 11, or 39.2
per cent., developed tumors. Eight of these subsequently were
absorbed. Microscopical appearances of this generation were simi-
lar to the original tumor.
9. Dark, gray, female mouse (Original xix), among the stock
animals of the institute, showed a tumor 1.3 centimeters in diameter,
situated beneath the skin of the lower portion of the abdomen, to
the left of the middle line. It had slightly invaded the muscles
of the thigh. The surface was irregular. The tissue was hemor-
rhagic and there Avere numerous cysts. In appearance it was found
to be composed of niunerous cysts and necrotic areas containing
hemorrhages, and areas of epithelial cells showing little definite
arrangement. Mitotic figures were common. The blood vessels
were greatly dilated.
From this tumor 20 mice, 10 of the color of the original, and
10 white, were inoculated. Of the 10 white mice inoculated, 2
died in a few days, and of the 8 surviving, one developed a tumor
after 45 days, which later became absorbed. One of the ten colored
mice developed a tumor after 120 days which was used for further
inoculation. Under the microscope this tumor was less adenoma-
tous in structure than the original. It showed hemorrhages and
necrotic areas. From this tumor 15 mice (colored?) were inocu-
lated, of which 2, or 13.3 per cent., developed tumors. In one the
incubation period was 160 days and in the other, 90. The structure
remained about the same as in the preceding generation. Further
transplantation was not undertaken with this tumor.
Digitized by
Google
J. TT. Jdbling. 101
lo. White, apparently old, female mouse (Original xx) showed
on neck anterior to the right of the median line, a tumor two and
one-half centimeters in diameter. The surface was gray and mot-
tled with dark brown, the latter portions being cystic. Under the
microscope this tumor was adenomatous in structure, consisting of
acini intermingled with solid masses of epithelial cells. Areas of
necrosis and hemorrhage occurred, the former being more frequent.
The acini were frequently dilated into cysts, some of which con-
tained a homogeneous eosin-staining material, and others, blood.
There were numerous mitotic bodies. The blood vessels were
widely dilated.
50 mice were inoculated, of which 5, or 10 per cent., developed
tumors. One of these later underwent absorption. The incubation
period was 30 days. The structure of these tumors differed from
the original in that the stroma was far better developed and was
very cellular (Plate XXIV, Fig. 16). Large hemorrhagic areas
occurred and the acinous arrangement had almost completely disap-
peared. There were large lobules composed of central cavities con-
taining cells, and hemorrhages were common.
40 mice were inoculated from this generation, of which 3, or 1.5
per cent., developed tumors, all of which underwent absorption
before they reached a size large enough to be used for further trans-
plantation.
CYST-ADENOMA.
Of this class there were two tumors, and since they occurred in
mice in which the other classes of tumors were present, and which
were of larger size, they were not submitted to transplantation.
1. The first of these, which occurred in Original ii, was situated
on the right side of the thorax and measured .6 centimeter in its
greatest diameter. In form it was flat and freely movable. The
tumor with which it was associated was a hemorrhagic cyst-
adenoma.
2. The second of these tumors was in Original v. It was situated
on the left side of the thorax, was flat, and measured 0.5 centimeter
in its greatest diameter. The tumor with which this was asso-
ciated belonged to the class termed by Apolant solid carcinoma.
Digitized by
Google
102 Spontaneous Tumors of the Mouse.
In histological structure these growths were very similar to one
another. They were made up of cysts developed from dilated acini,
and lined by a single layer of cells, some of which were flattened.
The cysts, however, frequently contained a homogeneous, eosin-
staining material. No mitotic figures were discovered. The
stroma was delicate in character.
The possibility, of course, exists that these second tumors of
smaller size were metastases. The type, however, of the smaller
tumors was wholly different from the larger ones, and approached
more nearly the typical adenomatous form than did the others.
Since each was distinctly incapsulated, and there were no mitotic
figures discovered, and no evidences available of growths having
developed in the lymphatic glands, they w^ere regarded as inde-
pendent tumors.
ALVEOLAR CARCINOMA.
According to Apolant, in this type of tumor, no acini, strictly
speaking, are present, but the tumor consists of lobules containing
alveoli closely packed with epithelial cells. One such tumor came
under observation among our series. It occurred in a white female
mouse (Original v), was situated in the left inguinal region and
measured 2 by 2 centimeters. Neither the skin nor the abdom-
inal muscles were involved in it. Its central portion was necrotic,
consisting of a cheesy mass, while the cut surface presented a gran-
ular appearance. Its external contours w^ere irregularly lobulated,
and nutrient blood vessels radiated over its surface. Near the left
axilla there was a flat nodular growth measuring 0.5 centimeter,
and presenting the same appearances as the large tumor.
Microscopically this tumor consisted of lobules filled with closely
packed epithelial cells, many of which through pressure had become
flattened and spindle-shaped. These solid masses of cells bore a
certain resemblance to spindle-cell sarcoma. Where the lobules
were large, the central portions were necrotic. Mitotic figures were
numerous. It should be stated that at one point where the stroma
was unusually w^ell developed a few acini occurred, and that the
second flat nodules presented the characters of cyst-adenoma.
49 mice w^ere inoculated from the alveolar tumor, of which 10,
Digitized by
Google
/. TT. Jobling. 103
or 20 per cent., developed tumors. The average incubation period
was 70 days. This tumor is now in the sixth generation, and the
highest percentage of implantations thus far obtained has been 59
per cent. The structure of the transplanted tumors is similar to
that of the original.
MOLLUSCOIDAL TUMOR.
A timior differing greatly from all others which have come to
our attention will now be described. It resembles closely a
tumor described by Haaland,*® and, owing to a certain resemblance
to the lesions of mollnscum contagiosum, he termed it "tumeur
molluscoide." He describes the tumor as being composed of lobules
with rounded bases at the periphery which radiate from a common
centre in the manner of the spokes of a wheel. The lobules are
broader at the periphery, and the central portions of them show
frequently the presence of keratin. The tumor which came to our
attention agrees closely with this one, although the degree of kera-
tinization was somewhat less. According to Haaland's view, the
tumor developed from hair follicles, and we are of the opinion that
our tumor took its origin in the same structures. He inoculated
twelve mice, but in no case did he get a secondary growth.
This tumor occurred in a full grown, brown, female mouse (Orig-
inal xxii). The growth was situated in the right inguinal region,
and measured 2 by 3 centimeters. It had become attached to
the skin, and ulceration was beginning. The tumor also penetrated
the abdominal muscles and projected slightly into the abdominal
cavity. It was grayish-white in color, rather firm, and showed one
hemorrhagic area. There were no cysts. In the upper portion of
the lower lobe of the left lung a nodule measuring one millimeter
in diameter was present.
At the margin of the tumor, where it invaded the neighboring
tissues, the cells were arranged in columns, consisting of solid growths
of epithelial cells (Plate XXV, Fig. 17). Where these columns
were cut transversely, they presented somewhat the appearance of
tubules. The general tendency of the tumor was to grow in such
columns. Tubular formations were also present, some of which
**Haaland, Ann, de VInst, Pasteur, 1905, xix, 163.
Digitized by
Google
104 Spontaneous Tumors of the Mouse.
branched and appeared to end blindly in solid masses of cells.
These tubules presented several layers of cells. Mitotic figures
were very common. Throughout the section irregular masses of
kerato-hyalin occurred ( Plate XXV, Fig. i8 ) . They usually occupied
the central portions of the solid lobules which were found both in
the periphery and central portions of the tumor. As a rule, these
areas were more abundant in the solid strands themselves, but they
also occurred in the tubular formations, the lumen of some of which
were filled with it. The nodule in the lung proved to be a primary
cyst-adenoma.
SO mice were inoculated with the large tumor, but none developed
secondary tumors.
PAPILLARY CYST-ADENOMA OF THE LUNG.
The class of tumors to which the primary lung tumors to be
recorded belong was first described by Livingood,** who found it
accidentally in the lung of a mouse dying of an experimental bac-
terial infection. Much later Haaland*^ reported five instances of
this tumor, but the most complete description in the literature is that
of Tyzzer,*® who described a series of twelve such tumors.*'^ It is
noteworthy that in Tyzzer's series of cases, six of the animals pre-
sented tumors in other parts of the body, or in other words, were
the subject of multiple primary tumors. Moreover, the circum-
stance should be mentioned that in the examples described by
Haaland the mice had previously been inoculated with tumor frag-
ments, so that he has considered the possibility of their origin from
the inoculated material. Murray*® has recently reported finding
this type of lung tumor, but he does not state with what frequency.
Our observations are limited to nine primary tumors of the lung
*• Livingood, L. E., Johns Hopkins Hosp. Bull, 1896, vii, 177.
* Haaland, Ann. de VInst. Pasteur, 1905, xix, 165.
**Tyzzer, E., Fourth Report of the Caroline Brewer Croft Fund Cancer
Commission, 1907, 27.
*' In a later report {op. cit.) Tyzzer states that he has met with 37 mice show-
ing primary lung tumors. In one family of 25 mice, four examples of primary
adeno-carcinoma of the lung were found; and in another family of 100 mice,
springing from a mouse with an adenoma of the lung, 15 mice developed tumors.
*' Murray, Scientific Report of the Imperial Cancer Research Fund, 1908, No.
3, 69.
Digitized by
Google
J. W. Jobling. 106
of this type, one animal of this series presenting two such tumors.
These tumors were distributed as follows : two of them occurred in
mice in which there were hemorrhagic cyst-adenomata, four in mice
with adeno-carcinomata, one in a mouse with multiple superficial
timiors, and one in a mouse in which the superficial tumor was a
spindle-cell sarcoma. It was this last animal which showed two
primary lung tumors, one in each lung. With one exception, the
tumors were all too small to be used for purposes of transplantation.
1. The first of these tumors which -we came across was in Orig-
inal Mouse xii. It was located in the lower margin of the left lung,
was grayish-white in color, and measured three millimeters in diam-
eter. It projected slightly above the surface of the pleura, and was
sharply outlined. Its microscopical appearance consisted of irregu-
larly projecting folds of connective tissue containing cavities which
were lined by columnar epithelial cells, the outlines of which were
sharp ( Plate XXVI, Fig. 19). A small bronchus, the lumen of which
was almost occluded by the growth, approached the periphery of
the tumor. The epithelial cells of the tumor and the bronchus were
almost indistinguishable. The central part had begun to degen-
erate and contained a greater amount of fibrous tissue. The cavities
of the tumor contained large cuboidal cells, undergoing degenera-
tion. Mitotic figures were not discovered. The pulmonary tissue
immediately adjacent to the tumor was collapsed. The epithelial
cells covering the papillary outgrowths of connective tissue formed
a single layer, and the stroma was small in amount.
Serial sections showed the growth to be wedge-shaped, and the
apex to consist of two or three papillary outgrowths from the wall
of a bronchiole. From the front of the organ the growth extended
along a bronchiole into the larger bronchi and the surrounding tissue.
2. The second tumor of this type occurred in Original Mouse x.
The nodule measured 0.5 millimeter and was located in the upper
lobe of the left lung, projecting slightly above the surface of the
organ, and being sharply outlined. Under the microscope this tumor
proved to be more solid than No. i. It also invaded the pulmonary
alveoli to a larger extent, which it appropriated to supply its frame-
work. The individual cells were round and the nuclei were rich in
chromatin. Mitotic figures were not discovered. In the spaces
Digitized by
Google
106 Spontaneous Tumors of the Mouse.
between the papillary outgrowths numerous rather large cuboidal
cells were present.
3. The third tumor w^as found in Original Mouse xiii. It was
not seen at the original examination of the animal, and was detected
subsequently in the study of sections of the lungs. The tumor was
located in the central portion of the upper lobe of the right lung.
In structure it resembled lung tumor No. i, except for the epithelial
cells, which were cubical in shape ( Plate XXVI, Fig. 20) . The general
structure of the tumor was also more compact, since the spaces
between the papillary projections were smaller. Serial sections
showed the tumor to be conical in shape, and the apex of the cone
to lie in close relation with and apparently to spring from a terminal
bronchiole (Plate XXVII, Fig. 21).
4. This tumor was found in Original Mouse xvi. On the anterior
surface of the lower lobe of the left lung, a grayish-white mass one
millimeter in diameter existed. It projected above the pleura and
was sharply outlined. Under the microscope this mass was found
to be composed of papillary outgrowths covered by a single layer
of epithelial cells of cubical form. The cells were smaller than in
the preceding tumors of this type. A special feature of this tumor
was the presence of a homogeneous material separating the papillary
projections in the more solid portions, and filling the small cystic
spaces. Mitotic figures were not discovered. There was compres-
sion of the surrounding lung tissue.
5. The next tumor of this class was found in Original Mouse xx.
It was situated in the upper lobe of the right lung and measured
one and one-half centimeters in diameter. Its structure was
very similar to those already described. The central portion was
fibrous, because of degeneration of that part of the tumor. This
fibrous tissue contained clefts in w^hich there had been fatty acid
crystals, and the clefts were surrounded by foreign body giant cells.
The surrounding lung tissue w^as collapsed.
6. The next tumor of this type occurred in Original Mouse xxii.
It measured one millimeter in size and occurred in the upper portion
of the lower lobe of the left lung. It resembled in structure the
previous tumors described.
7. The seventh of the series occurred in Original Mouse xxv. In
Digitized by
Google
J. W. Jobling. 107
this animal the lung showed several nodules, two of which were
situated near the apex of the right lung, and measured about two
millimeters in diameter. One of these only was found to be a
primary pulmonary tumor, the others being metastases of the super-
ficial adeno-carcinoma. The primary tumor corresponded to the
general tumors of this type and differed greatly from the metastatic
nodules.
8. In Original Mouse xxvi two primary lung tumors were present.
One was found in the upper lobe of the right lung and the other in
the upper lobe of the left lung. They measured about one milli-
meter in diameter. The superficial tumor with which they were
associated was a spindle-cell sarcoma. In structure they were quite
similar to others of this class (Fig. 12).
SARCOMA.
The literature does not contain, as far as I have been able to see,
records of the successful transplantation of primary sarcoma of the
mouse into other mice, although sarcomata of rats have been in
several instances transplanted successfully. Ehrlich and Apolant*^
have reported the finding of two mixed tumors composed of sar-
comatous and carcinomatous elements in the first generation, and
Murray^^ alludes to an instance of a spindle-cell sarcoma found by
Jensen in a mouse sent to him by Bashford.^^ In the latter instance
the growth surrounded one of the kidneys and although it was trans-
planted into a large number of mice, it gave rise to no implantation
tumors. In the instances mentioned as having been reported by
Ehrlich and Apolant, they regard the tumors as having been original
carcinomata, which were being transformed during the original gen-
eration into sarcomata. Ehrlich and Apolant,*^^ Loeb,^^ Liepman^*
*• Ehrlich and Apolant, Berliner klin. Woch., 1907, xliv, 1397.
" Murray, Scientific Report of the Imperial Cancer Research Fund, 1908, No.
3, 73-
" Bashford and Murray, in a later article (^Proc. of the Royal Med. and Chir.
Soc. of London, Series B, 1909, Ixxi, 311), mention a melanoma or melandric
sarcoma of the ear of a mouse, but do not allude to transplantation.
"Ehrlich and Apolant, Berliner klin, Woch., 1905, xlii, 871.
"•Loeb, Med. Bull, 1906, xix, 113.
•• Liepman, MUnchener tned. Woch., 1907, liv, 1345.
Digitized by
Google
108 Spontaneous Tumors of the Mouse.
and Bashford^^ all report examples of the transformation of car-
cinoma of the mouse into spindle-cell sarcoma. In all of these
instances excepting one, namely, that of Loeb, the changes occurred
in relatively late generations, whereas the example reported by Loeb
is the first instance in which the change began in the first generation.
I have observed two instances of what appeared to be primary sar-
comatous superficial tumors of the mouse. In both instances tissue
was examined from all parts of the tumors in order to exclude as
far as might be possible a mixed carcinoma and sarcoma in the first
generation. No evidence of epithelium or of cancerous tissue was
detected.
1. The first of these tumors was found in a white, adult, female
mouse (Original xviii) . It was located on the left side immediately
inferior to the foreleg, and extended over the lower ribs. It meas-
ured three centimeters in diameter. The skin was invaded, and
there was an ulcer measuring half a centimeter in diameter
which was superficial and covered with a scab. The tumor was
grayish-white in color, firm in consistence and had extended into the
muscles about the ribs and projected into the pleural cavity as a flat
mass with a concave surface. On section it presented central necrotic
masses. There were no metastases, but both ovaries were enlarged,
firm, oblong in shape and about eight times their normal size.
Under the microscope this tumor proved to be a spindle cell sar-
coma, possessing highly invasive properties and insinuating itself
between the voluntary muscle fibers (Plate XXVII, Fig. 22). The
ribs were surrounded by tumor cells, but were not invaded them-
selves. The enlarged ovaries were the seats of papillary cyst-adeno-
mata, which will be described in another place.
60 mice were inoculated from the superficial tumor, of which 45
survived two weeks or longer. In one instance a tumor developed,
which was detected on about the 70th day. The tumor grew slowly
and a portion of it was removed for examination with a view of
stimulating the growth of the remainder. The mouse succumbed
soon after the operation, before the growth had reached a size large
enough for further transplantation.
2. This tumor occurred in Original Mouse xxvi. It was situated
" Bashford, Berliner klin. Woch., 1907, xliv, 1238.
Digitized by
Google
J. W. JoUing, 109
on the left side of the chest just below the fore leg, measured three
centimeters in diameter, was flattened and had invaded the
skin. A small ulcer had appeared which was also attached to the
chest wall, although it had not penetrated to the pleura. It was
grayish-white in color and rather firm in consistence. Tissues
immediately surrounding the ulcer were hemorrhagic, while around
the main tumor they were edematous. A small nodule measuring
one millimeter in diameter was found in the lower portion of the
upper lobe of the right lung. The corresponding portion of the
left lung showed a similar nodule. These were primary tumors
which have been described elsewhere. Both ovaries were enlarged
to about five times the normal size.
The microscope showed this tumor to be also a spindle cell sar-
coma which had invaded the adjacent muscle fibers. Many of these,
or remains of them, were contained in the mass (Plate XXVIII,
Fig- 23). A large number of eosinophilic cells were scattered
through the tumor. The left ovary was the seat of a cyst adenoma
to be described elsewhere.
40 white mice were inoculated from the superficial tumor, of
which 14, or 35 per cent., developed tumors. The average incuba-
tion period was 30 days. The tumors presented the same histo- .
logical character as the original, and are now growing in the second
generation.
LYMPHO-SARCOMA AND HODGKIN'S DISEASE.
The first tumors of this class were described by Haaland.^® He
described in all six examples of mice in which the lymphoid tissue
throughout the body was so greatly increased that he concluded that
the condition consisted not of tumor formation in the true sense
so much as an hypertrophy or hyperplasia of the preexisting lym-
phoid structures. He inoculated other mice with the lymphatic
material without results. On the other hand, through having kept
mice in a cage with an animal affected in this way, he observed five
similar formations in the mice introduced into the cage within
a period of two years. Tyzzer*^*^ reports two instances of a condi-
"• Haaland, Ann. de VInst. Pasteur, 1905, xix, 197.
"Tyzzer, Fourth Report of the Caroline Brewer Croft Fund Cancer Com-
mission, 1907, 27.
Digitized by
Google
110 Spontaneous Tumors of the Mouse,
tion somewhat resembling the condition described by Haaland, but
he states that they presented the appearance of a "well-defined
primary growth without appreciable hyperplasia of the lymphoid
tissues elsewhere in the body."^® Murray*^® reports four similar
instances. In two, a general hyperplasia of the lymphoid tissue
existed, and in two, the tumors were more localized. Both Tyzzier
and Murray failed to transplant successfully the tumors to other
mice.
I. The first example which I observed occurred in a white, female
mouse (Original xvii). On each side of the neck were masses
measuring 1.2 centimeters in diameter which were freely movable
and not attached either to the skin or to the deeper tissues.
The axillae contained masses measuring from five to seven milli-
meters and the right inguinal region, a mass measuring five milli-
meters. All of these presented the same naked eye appearances.
They were grayish-white in color, quite uniform and not lobulated.
The liver was mottled, owing to the presence on the usual dark
ground of many minute grayish points. The upper pole of the
right kidney contained a grayish nodule measuring one and one-half
millimeters and the left kidney showed two smaller nodules. Two
small masses occurred in the retro-peritoneal tissues just behind the
left kidney. Two masses measuring seven millimeters in diameter
were attached to the mesentery. The lungs also contained several
small foci, gray in color and sharply outlined. The spleen was
considerably enlarged and mottled.
Under the microscope the structure of all of the different masses
described proved to be the same. The masses were made up almost
exclusively of small round lymphoid cells possessing a compara-
tively large pale vesicular nucleus and a narrow rim of cytoplasm.
The cells were of the type found in the germinal centres of the
lymph follicles. The stroma was delicate except where the larger
blood vessels coursed. Mitotic figures were not numerous. In
addition to the larger cells described, others of the usual lympho-
"Tjrzzer states in his later paper (op. cit.) that he has met with 10 mice
showing lympho-sarcoma ; and Bashford and Murray in their last report {op.
cit.) record another instance of generalized lympho-sarcoma.
" Murray, Scientific Report of the Imperial Cancer Research Fund, 1908,
No. 3, 74.
Digitized by
Google
J. W. Johling. Ill
cytic type were present. Small necrotic areas occurred here and
there in some of the nodules.
The lungs showed much more replacement of the pulmonary
tissue by the lymphoid growth than was noticeable to the naked eye.
When the lymphatic nodules were small they were confined to the
perivascular and peribronchial tissues, but the larger masses also
invaded the pulmonary tissue. In addition, small accumulations of
cells of the same general type occurred beneath the visceral pleura.
Mitotic figures were fairly common. In other respects the lungs
were congested and edematous, and contained numerous phagocytic
cells containing pigment. It may be remarked that the pulmonary
condition was similar to that described by Haaland.
The spleen was greatly modified, since both the pulp and the
follicles were diffusely infiltrated wath the lymphoid cells in such a
way as to make the normal structure of the organ indistinct.
Many myeloid giant cells were present, as is not uncommon in the
spleen of the normal mouse. The sinusoids of the spleen were
greatly distended with the cells, which, on the whole, appeared
larger than the average. Mitotic figures were numerous.
The liver also exhibited an extensive invasion with the new cells.
In some places the growth of these cells had replaced the normal
tissue to such an extent that nothing of the liver structure remained
visible. In other places the cells had grown into and distended the
sinusoids, producing a compression of the rows of liver cells.
Sometimes the new cells had spread out diffusely and at other times
they had been confined to definite Ibculi. The interlobular tissue
was more affected than other parts.
The new cells in the kidney were confined almost entirely to the
cortex of the organ. The growth in the cortex was intertubular
and periglomerular ; in the medulla it surrounded the vessels only.
I GO mice were inoculated with the material from the superficial
masses, 70 were inoculated from the masses in the neck, 10 from
the mass in the left axilla, 10 from the mass beneath the right pec-
toral muscles, and 10 from the mass in the right inguinal region.
At the end of two weeks 60 of these mice remained alive, of which
one developed a tumor. This growth was first noticed 120 days
after the inoculation. When the growth had attained the size of
Digitized by
Google
112 Spontaneous Tumors of the Mouse.
four by seven millimeters, a portion was removed for further inocu-
lation and for histological study. The fragment which was left
behind underwent absorption. 20 mice were inoculated with the
excised fragment, but none developed tumors. Under the micro-
scope the structure of this implantation nodule was in all respects
similar to the original growths, except that the mitotic figures were
more numerous, as were the capillaries. The stroma presented the
character of the stroma of the original nodules.
2. The second example appeared in a white, female mouse (Orig-
inal xxi) which showed a superficial tumor in the right inguinal
region. The mass measured one centimeter and was freely
movable. A portion of the tumor was removed by operation. It
was grayish-white in color and somewhat lobulated. The fragment
of tumor left behind did not increase appreciably in the next 14
days, the period during which the mouse survived. The autopsy
showed a small mass one millimeter in diameter in the upper margin
of the lower lobe of the right lung. It appeared to be a metastasis.
Unfortunately the tissues removed at the autopsy were lost, and only
the portions removed by operation remained for further study.
This tumor was composed for the most part of cells of the lym-
phoid type which were somewhat smaller in size than in the pre-
ceding case. While many of the cells resembled the lymphocyte,
others were somewhat larger and possessed pale vesicular nuclei.
The stroma consisted of a delicate reticulum enclosing small masses
of cells. In the periphery there were many capillaries. Mitotic
figures were not very numerous. Two appearances are highly
important : in the first place there were many eosinophilic cells scat-
tered through the tissue, and next there also occurred here and
there, and particularly in the more central parts of the growth, large
areas composed of a reticulum coarser than the usual reticulum and
containing spaces enclosing large endothelioid cells, of which some
were multinuclear. In addition these areas contained large and
small lymphoid cells, a smaller number of eosinophilic and a large
number of plasma cells. The endothelioid cells presented irregular
margins and contained a large amount of cytoplasm that was strongly
eosinophilic. The nuclei of these cells were either vesicular or
solid and dense. A certain number of mitotic figures were present
Digitized by
Google
J. W. Jobling. 113
among these cells. The large multinucleated or giant cells showed
nuclei which were usually vesicular and contained a coarse chro-
matin network which was disposed in an irregular fashion in the
cytoplasm. In these areas the l3rmphocytes were few in number,
and the other cells relatively numerous. A small number of cells
containing basophilic granules were also present.
The description just given indicates that the tumor in this animal
possessed the histological characters of Hodgkin's disease, as has
been described by Reed, Longcope and others, in human beings. It
is true that the eosinophilic cells were perhaps less numerous in the
mouse nodule than is the case usually in the human disease, but
Longcope has pointed out that in the later stages of the disease in
human beings the eosinophilic cells also diminish. Whether, there-
fore, we are to regard the nodule in the mouse as representing a late
stage of the disease, or whether in this animal the eosinophilic cells
tend to be less numerous, must be left undecided.
OVARIAN TUMOR.
I have observed two examples of tumors affecting the ovaries,
and it is interesting to note that in both cases the large superficial
tumors also present consisted of spindle-cell sarcomata.
I. The first example occurred in Original xviii. In this white
mouse both ovaries were enlarged, firm in consistence, oblong in
shape, and about eight times the normal size. Under the micro-
scope the two organs presented the same appearances which were
interpreted as representing papillary cyst-adenomata. There had
taken place a great increase of the epithelial cells, which formed
solid masses somewhat compressed and elongated into spindle cells
and cysts of different sizes, usually small and more or less occupied
by the papillary outgrowths from their walls. These outgrowths
developed from narrow or somewhat thicker pedicles and spread
out into a fan-like structure. The epithelial cells which covered
the papillae and the walls of the cysts were rich in cytoplasm and the
nuclei of the cells were usually vesicular but sometimes dense. The
more solid portions of the organs were at one time cystic but the
cysts became occluded by the ingrowth of the papillae. Acini
Digitized by
Google
114 Spontaneous Tumors of the Mouse,
possessing a distinctly granular form and arrangement also occurred.
Mitotic figures were rarely seen, but appearances suggestive of
direct cell division were more frequent. Hemorrhage had taken
place into some of the cysts. Sections stained by Mallory's con-
nective tissue method revealed a delicate basement membrane of
connective tissue surrounding the small cysts and the more solid
portions as subdivided into small areas of quite definite form.
These latter areas were filled with epithelial cells of a granular
quality resembling somewhat the lutein cells of the ovary.
2. The second example of ovarian tumor occurred in Original
Mouse xxvi. The left ovary was about five times the normal size
and showed under the microscope numerous large cysts separated
by smaller ones and the tubules of the ovary. The cysts and tubules
were lined by high columnar epithelium, but cilia were not demon-
strated. In some of the larger cysts the lining epithelium had
become flattened. The tissue separating the cysts was composed to
a large extent of smooth muscle cells arranged in strands or in
bundles. Sections stained by Mallory's phosphotungstic acid and
hematoxylin brought out the myoglia fibrils. The nuclei of these
muscle cells were large and vesicular and showed a few mitotic
figures. In certain portions the granular structures were few and
the muscular fibres many, so that the appearance presented was that
of a leiomyoma.
CONCLUSIONS.
There have come to our hands within a period of about two years
and as a result of relatively non-strenuous efforts 26 mice which
had developed spontaneous tumors, and according to our interpre-
tations these 26 mice were the subjects of 41 primary tumors.
Our experience agrees with the observations of others that it is
especially, and perhaps almost exclusively, the female mice which
develop spontaneous tumors. In no instance has a male mouse pos-
sessing a spontaneous tumor come into our hands.
Our experience is further in agreement with previous observa-
tions which teach that by far the greater number of the spontaneous
tumors of the mouse develop from the mammary glands. Of the
41 tumors which we have described, 29 arose in portions of the body
corresponding with the distribution of the mammary glands.
Digitized by
Google
/. W. Jobling. 115
Choosing the classification adopted by Apolant, we have found
among the true tumors of the mammary glands representatives of
each of his classes.
We also observed an example of an adenomatous tumor arising
from a portion of the body corresponding with the distribution of
the mammary gland which exhibited the structure of the typical
sebaceous glands rather than the mammary gland, and which we
have classified under the moUuscoidal tumors of Haaland.
Metastases occurred in a certain number of the mice showing
spontaneous tumors. When present they were exclusively in the
lungs. We did not observe a single instance of metastasis to other
organs, or to the lymphatic glands.
Next to the mammary glands the lungs showed in our series the
most numerous primary tumors. According to our interpretations
we encountered nine examples of primary pulmonary tumors, which
were always associated with superficial tumors of the body. It is
noteworthy that we have not come across a single instance of a
primary pulmonary tumor among several thousand mice not show-
ing superficial spontaneous tumors, which have been subjected to
careful post-mortem examination. It would be easy to conclude,
therefore, that the occurrence of spontaneous tumors in one part
of the body in mice predisposes them to the development of spon-
taneous tumors of the lungs, but obviously such a deduction is not
warranted at present. The first example of this type of tumor was
described by Livingood.
Our observations of two primary sarcomata of the superficial
tissues of the mouse is apparently unique. We believe that we
excluded, probably, the possibility that the tumors were mixed sar-
coma and carcinoma. Moreover, they exhibited properties unlike
the superficial carcinoma of the mouse, since they were far more
invasive locally. Our success in transplanting both of these sarco-
mata is also unique. In the first case a single successful graft which
was not further transplanted was obtained ; in the second case the
sarcoma is being successfully transplanted at the present time.
We observed two examples of tumors developing from the lym-
phatic organs. One of these belonged obviously to the so-called
lympho-sarcoma of mice. The other, we believe, is unique in so far
Digitized by
Google
116 Spontaneous Tumors of the Mouse.
as Its identification with the pathological condition described in
human beings under the name of Hodgkin's disease is concerned.
Furthermore, we succeeded for the first time, apparently, so far as
the records in the literature go, in transplanting successfully a
lympho-sarcoma. In one mouse an undoubted development of the
transplanted graft occurred.
We also observed two examples of ovarian tumors, both of which
were of the nature of papillary cyst-adenoma. These tumors arose
in mice which were the subjects of spindle-cell sarcoma of the super-
ficial tissues.
Our experience in the transplantation of the spontaneous mouse
tumors agrees generally with that of other investigators. It would,
however, appear that we have been on the whole more successful
than others in the transplantation of the hemorrhagic class of mam-
mary gland tumors, the so-called hemorrhagic cyst-adenomata. An
explanation for this discrepancy cannot be readily given. It is, of
course, possible, since such great variations, are known to exist
among mice as regards their susceptibility to tumor implantation,
that the American mice are more subject to implantation of the
hemorrhagic tumors than the European mice, but it is also possible
that we have interpreted somewhat more broadly than others the
class of hemorrhagic tumors. If the latter statement is true, the
tendency was an unconscious one, and suggests that different work-
ers will probably interpret these tumors quite differently.
Finally, this study indicates that mice are subject to a wide
variety of spontaneous tumors, and suggests that the more widely
the study is carried, the greater will be found the number and variety
of tumors to which they are susceptible.
For the present we prefer to believe that spontaneous tumors are
more frequent in female than in male mice, but that they also will
be found to occur perhaps not very uncommonly among the latter.
The point should be borne in mind that by far the greater number
of spontaneous ttimors of mice which have thus far come under
observation have been present in old female mice. The business of
breeding mice for sale leads to the retention of the female breeders
until they become exhausted in this function by age, and the chang-
ing and weeding out of the males while they are still much younger.
Digitized by
Google
MONOGRAPH NO. 1. PLATE XVII.
Fig. I.
Fig. 2.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XVIII.
Fig. 3.
Fig. 4.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XIX.
Fig. 5.
Fig. &.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XX.
Fig. 7.
Fig. 8.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XXI.
Fig. 9.
Fig. 10.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XXI!.
Fig. II.
Fig. 12.
Digitized by VjOOQIC
Digitized by
Google
MONOGRAPH NO. 1. PLATE XXIII.
Fig. 13.
Fig. 14.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XXIX.
Fig. 15.
Fig. 16.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XXV.
VlG. 17.
Fig. 1 8.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XXVI.
Fig. 19.
Fk;. 20.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XXVII.
Fig. 21.
Fig. 22.
Digitized by
Google
Digitized by
Google
MONOGRAPH NO. 1. PLATE XXVIII.
Fig. 23.
Digitized by
Google
Digitized by
Google
J. W. JohUng. 117
Consequently, far fewer males reach comparatively old age in the
sense that the female breeders do, and this may be an important
factor in accounting for the difference in susceptibility to spon-
taneous tumors hitherto observed in the two sexes.
EXPLANATION OF PLATES.
Plate XVII.
Fig. I. Original Mouse Tumor iv. Adeno-^arcinotna. In the central and upper
portions the growth is almost purely adenomatous, although there are also
several nests of cells in which the glandular arrangement is not evident. The
lower right hand portion shows the solid areas more clearly.
Fig. 2. Daughter tumor of fifth generation. The growth here is solid and
there are no evidences of adni. The stroma is edematous. The presence of fat
cells just below the center and to the left of it denotes the invasive character of
the tumor.
Plate XVIII.
Fig. 3. Daughter tumor of sixth generation. Shows apparent reversion to
original type. It is con^osed almost wholly of acini and shows marked edema
throughout It is very similar to the primary tumor, though the stroma is not
so well developed.
Fig. 4. Primary Tumor vi. The tumor is composed mainly of acini, some of
which have become dilated and form small cysts. The stroma is not well
developed. Solid nests of cells are not infrequent.
Plate XIX.
Fig. 5. Daughter tumor of second generation. The growth is solid and
there are no evidences of acini. The stroma is very delicate in character. A
few blood-vessels with very thin walls are to be seen.
Fig. 6. Daughter tumor of the second generation. Lobules show necrotic
centers. The stroma separating the lobules is well developed and very cellular.
Numerous large blood-vessels with very thin walls are to be seen. The appear-
ance here is very suggestive of a beginning sarcoma.
Plate XX.
Fig. 7. Primary Tumor xiii. Adeno-carcinoma, The growth is composed of
acini which are separated from one another by a very deficate stroma. Several
blood-vessels are to be seen.
Fig. 8. Daughter tumor of third generation. The growth is composed of
lobules, the central portions of which are necrotic. A few acini are to be seen
between the lobules.
Plate XXI.
Fig. 9. Primary Tumor xxiv. Adeno-carcinoma. This shows a small area
of keratinization in the more solid portion of the tumor. Numerous acini are
present
Digitized by
Google
118 Spontaneous Tumors of the Mouse.
Fic. 10. Primary Tumor xxv. The mass of keratin in the center is sur-
rounded by numerous large cysts and a number of normal appearing acini. The
cysts are filled with a finely granular and faintly staining material. In places the
acini are widely separated by the serous exudate.
Plate XXII.
Fig. II. Primary Tumor xxv. Extra vascular lung metastasis. There are
numerous acini, although the growth is composed mainly of nests of cells.
Several mitotic figures are present. The stroma is very delicate.
Fig. 12. Primary papillary cyst-adenoma from the lung of the same mouse.
This figure is to be compared with Fig. ii. The former is a metastasis from
the large subcutaneous tumor shown in Fig. lo. The irregular papillary in-
growths are covered by a single layer of columnar epithelium. A few necrotic
cells are to be seen in the cavities. The connective tissue framewoilc is not
well developed. The different characters of the growths shown in Figs, ii and
12 justify us, we think, in concluding that the tumors are entirely independent
of one another in their origin.
Plate XXIII.
Fig. 13. Mouse Tumor iii. Daughter tumor of first generation. This shows
very well the hemorrhagic type of adeno-carcinoma. There are numerous large
cysts into which hemorrhages have occurred. Those cysts not showing hemor-
rhagic contents are usually filled with a serous exudate. The growth between
the cysts is usually solid in character, though a few acini are to be seen.
Fig. 14. Mouse Tumor iii. Tumor of second generation. The growth shown
here is to be compared with that of the first generation as shown in Fig. 13.
The tendency here is towards the formation of irregular clefts and papillary
ingrowths. This is the " spalten bildendes " type of Apolant Hemorrhages are
much less frequent in this tumor. The stroma is rather cellular and well
developed in places.
Plate XXIV.
Fig. 15. Primary Tumor x. This is also typical of the hemorrhagic type of
adeno-carcinoma. In the center are two large blood-vessels in the midst of
necrotic tissue. The walls of the vessels are very thin and are surrounded with
a thin mantle of tumor cells. Hemorrhages are to be seen everywhere through-
out the section. Numerous acini are present.
Fig. 16. Mouse Tumor xx. Daughter tumor of first generation. This tumor
is composed of lobules, the central portions of which are necrotic* but the
remarkable feature of the growth is the cellular stroma separating the lobules.
In places the appearance of this tumor is very suggestive of a mixture of
sarcoma and carcinoma.
Plate XXV.
Fig. 17. Primary Tumor xxii. Molluscoidal type. This shows the radiating
columns of cells at the periphery. Some of the columns show a lumen on cross
section. The next figure shows that a number also contain masses of keratin.
The appearance presented in the deeper portions is that of an adenoma.
Digitized by
Google
J. W. Jelling. 119
Fig. 18. Primary Tumor xxii. Portion of same tumor which shows masses
of keratin. The keratin occurs not only in large masses but is also seen in
longitudinal sections of the smaller tubules. On the right the branching of the
tubules is very evident
Plate XXVI.
Fig. 19. Primary lung tumor in Tumor Mouse xix. The structure here is
typical of most of the primary lung tumors of our series. The papillary in-
growths are covered with a single layer of rather high columnar epithelium.
The stroma is delicate in character. The cell outlines are distinct and they can
easily be differentiated from one another.
Fig. ao. Original Mouse Tumor xiii. Primary lung tumor. This growth is
more compact than any of the others of this series. To the left of the center
there is a small bronchus, the lumen of which has been entirely occluded by the
tumor growth.
Plate XXVII.
Fig. 21. Original Mouse Tumor xiii. This shows what apparently is the
point of origin of the primary lung tumor shown in Fig. 20. The growth was
wedge-shaped and this corresponds with the apex which apparently was located
at the termination of one of the bronchioles.
Fig. 22. Primary Tumor xviii. Sarcoma. The growth is composed of spindle-
shaped cells which can be seen invading the surrounding muscle tissue. Some
of the muscle cells are undergoing pressure atrophy, but the majority are well
preserved.
Plate XXVIII.
Fig. 23. Primary Tumor xxvi. Sarcoma. The cells are irregular in size and
shape, but the dominant type is that of a spindle-celled sarcoma.
Digitized by
Google
TRANSPLANTATION EXPERIMENTS IN MACACUS
RHESUS WITH A CARCINOMATOUS TERA-
TOMA FROM MAN.
By J. W. JOBLING, M.D.
(From the Laboratories of the Rockefeller Institute for Medical Research,
New York,)
Although many attempts have been made to graft human tumors
upon the lower animals none have been really successful. The
experiment of Roux and MetchnikofP with anthropoids was a
failure. The use of the dog, rat, etc., has not given promising
results. The remarkable experience Lewin^ relates of the develop-
ment of a granulation tissue tumor in a dog inoculated from a
malignant adeno-cystoma of the ovary has not yet been satisfactorily
accounted for.
With our present knowledge gained chiefly from studies of trans-
plantable tumors of mice, rats and dogs, we can predicate that even
when transplantation is easily accomplished within species it fails
when attempted between nearly related species. Hence we should
not expect that fragments of tumors removed from human beings
should develop to any extent in other species of animals. However,
a final decision on this point may well be deferred until a number of
essays with a wide variety of tumors and several species of animals
have been made. At present we can view the problem of engrafting
tiunors upon heterologous species f roifi two points : first, in relation
to true grafting in which the tumor fragment actually takes hold
and grows ; and second, in relation to mere ability to increase some-
what in size for a short time, owing to temporarily favorable condi-
tions of nutrition and restraint of autolytic disintegration. This
indirect form of growth, which Ehrlich^ first described under the
* Roux and Metchnikoff, Bull, de L'Acad. de mid,, 1903, 1, loi.
* Lewin, Zeit f. Krebsforsch., 1906, iv, 55.
* Ehrlich, P., Arb. a. d. k. Inst. f. exper. Therapie, 1906, No. i, 2.
120
Digitized by
Google
J. W. J oiling. 121
name of zig-zag transplantation, does not always occur, and prob-
ably is dependent somewhat upon nearness of relationship between
the species — one that supplies and one that receives the grafts.
That some such factor operates at times is rendered probable by the
far greater reactions of inflammation which occur when the species
are widely separated than when nearly related. Dr. Flexner and I
have made observations of this kind with a transplantable rat tumor
ingrafted in mice, rabbits, and guinea-pigs. A similar interpreta-
tion can be put upon some of Loeb's* experiments.
It is now generally recognized that in order to secure satisfactory
results the transplantations must be carried out with great care so as
to avoid bacterial infection and consequent inflammatory reaction,
and no foreign fluids should be allowed to come into contact with
the graft. Unless these precautions are observed failure may result
even with readily transplantable tumors, and when the fate of the
grafts is to be followed day by day by accurate measurements, the
inflammatory reactions about the fragments must be controlled.
From this source alone actual increase in size may occur and it
can be excluded as a factor only by excising some of the enlarged
grafts and subjecting them to microscopical examination.
The history of the tumor, which follows in brief form, is con-
vincing as to its malignancy.
S. L., female, white, age 14 years. Patient was first admitted to the Pres-
byterian Hospital in the service of Dr. A. J. McCosh on January i8» 1908.
One sister when sixteen years of age had a dermoid cyst removed. Patient had
an enlargement of the abdomen which had been present for two months. A
diagnosis of a dermoid cyst of the ovary complicated by intestinal obstruction
was made. The patient was operated upon the day of admission. There was
found a smooth walled cyst the size of a foot-ball, which was friable and con-
tained a mass of heterogeneous tissue. The cyst was first drained and then
drawn into the incision, when a large amount of g^mous gelatinous material
escaped.
Pathological Report— {Kindly supplied by Dr. Opie.) The tumor is made
up of fibrous tissue, some striated muscle, and a number of hair follicles. There
are also some papillomatous structures and epithelial whorls.
The patient appeared relieved until twelve days after the operation, when
pain, vomiting and distention of the abdomen appeared, and with these symp-
toms a perceptible mass in the original locality. The mass continued to develop
until February 9, when it was again operated upon. At the operation there
*Loeb, L., Jour, of Med. Research, 1902, viii, i.
Digitized by
Google
122 Transplantation Experiments in Macacus Rhesus.
was found a mass of old organized and inorganized blood in the region of the
hepatic flexure of the colon and considerable fibrin and masses of tissue looking
like cyst elements (bone, cartilage, etc.) in the lower portion of the abdomen.
The report from the pathological laboratory on the material obtained at this
operation was "that it consists of masses of hyaline cartilage which are infil-
trated with a few polymorphonuclear cells. The stroma is composed of fibrous
tissue, fat and extravasated blood, and contains numerous cells of the poly-
morphonuclear and lymphoid type."
The patient was discharged three weeks later apparently well, but after
another three weeks returned on account of a growth about the size of a cocoa-
nut in the original locality. She was operated upon the following day, when
there was found a mass filling the right lower quadrant of the pelvis, which
was composed of a tenacious, hemorrhagic material involving all the structures
in this locality and having no apparent point of orig^in. The serosa of the
intestine contained a number of miliary nodules. It was with a portion of the
material removed at this operation that the first transplantation experiments
were conducted.
The material which was used for inoculation consisted of a soft,
friable mass of the consistency of brain tissue showing hemorrhagic
spots.
Microscopically the tissue consisted of hyaline cartilage, glandular
columnar epithelium and a cellular stroma. The stroma is rich in
cells and in places edematous. The cells are spindle-shaped or
polygonal with rather large, pale vesicular nuclei. In places the
stroma is finely granular and contains numerous fine vacuoles, this
appearance probably being due to edema. It is exceptional to find
a definitely fibrillated area, the tissue probably being too young for
this. Not infrequently one finds a small clump of embryonic
striated muscle cells in the stroma. Small hemorrhages are present.
Mitotic figures are fairly numerous. Imbedded in this stroma are
masses of hyaline cartilage. These masses are usually small and
almost round in shape, though in some instances they are larger and
more irregular in form. They are scattered throughout the sec-
tions and do not appear to bear any definite relation to either stroma
or clumps of epithelial cells. They are very cellular and many of
the cells are undergoing mitosis. Some of the masses are sharply
differentiated from the surrounding stroma by a layer of closely
packed cells, the nuclei of which are oval or elongated, with the long
axis pointing either toward, or parallel with the periphery. Mitotic
figures are present in the central as well as in the peripheral portions.
Digitized by
Google
J. W. Jobling. 123
The epithelial type of cell is present everywhere, either in the
form of irregular clumps without any definite arrangement, or in
the form of an acinar or gland-like arrangement lining irregular
clefts and tubules. In most instances the cells lining the tubules or
clefts are arranged in several layers, though, exceptionally, places
are seen in which the tubtdes are lined by a single layer. Goblet
cells are not infrequent among those lining the tubules. The cells
are usually large, with vesicular nuclei, though nuclei rich in chro-
matin are not infrequently seen. Without having regard to the
arrangement of the cells, mitotic figures are very numerous.
Owing to the rapidity of the growth after recurrence of the
tumor, Dr. McCosh offered it for inoculation experiments. We
obtained the tumor immediately after its excision, wrapped in sterile
dressings to prevent infection and evaporation, and made the
implantations into macacus monkeys, rats and mice. The technic
used was that followed in the transplantation of rat and mouse
tumors. The tumor was cut into small fragments, no fluid of any
kind being used for the subcutaneous inoculations, and one of the
fragments placed in the distal end of a hollow needle. After cut-
ting away the hair over the site of inoculation, the surface was
cleansed with alcohol and the alcohol removed with sterile gauze.
The needle was then introduced into the subcutaneous tissues and
by means of a sterile platinum wire, just fitting the lumen, the frag-
ment was pushed out of the needle into the tissues beyond. The
inoculations were always made beneath the skin of the abdomen.
Every precaution was taken to avoid infection. The time elapsing
between the removal of the tumor from the patient and its inocula-
tion into the animals was about forty-five minutes, the chief loss of
time arising from the distance between the Presbyterian Hospital
and the Rockefeller Institute.
Monkey J.— This animal was inoculated in eight places with the tumor frag-
ments. In four places it received tumor fragments alone ; in two places, tumor
fragments plus a drop of a saturated solution of scharlach R. in olive oil; in
two places, tumor fragments plus a drop of a saturated solution of sudan III
in olive oil ; in one place, scharlach R. solution atone ; and in one place sudan III
solution alone.
7th day. There are distinct masses to be felt at the points where the frag-
ments were lodged. These measured from 2 to 3 mm. in diameter.
Digitized by
Google
124 Transplantation Experiments in Macacus Rhesus,
9th day. Fragments about the same size. At this time the larger portion
of two of the fragments were excised. The fragments to which the scharlach
R. and sudan III solutions had been added at time of inoculation do not appear
as large as those inserted without any chemicals. The material excised was
used to inoculate two other monkeys.
nth day. The nodules not operated upon remain about the same size. Con-
siderable swelling at sites of operation.
13th day. The nodules are somewhat smaller than at previous examinations.
i6th day. Excised one of the nodules for histological study. Fragment
removed measured about 2 mm. in diameter.
i8th day. Fragments smaller, but still perceptible.
20th day. Fragments just perceptible.
22d day. Fragments disappeared.
The material excised from Monkey i was used to inoculate Monkeys 2 and 3.
Monkey 2, This monkey was inoculated in five places with material excised
from Monkey i.
4th day. All the fragments are easily felt and measure 2 mm. in diameter.
5th day. Excised fragment for histological study. The fragments measure
3 mm.
7th day. No change in size.
nth day. Fragments somewhat larger, measure 4 mm.
15th day. Fragments measure about 3 mm.
19th day. Fragments smaller.
21 St day. Fragments disappeared.
Monkey 3. — ^The animal was inoculated in five places with material excised
from Monkey i.
4th day. Fragments measure 2 mm. in diameter.
7th day. One of the fragments remains about the same size, the others
somewhat smaller.
I2th day. Fragments somewhat larger and measure 4 mm.
15th day. Fragments disappeared.
As Stated in the protocols, several of the fragments were removed
for histological examination.
The first fragment examined was excised from Monkey i on the
eleventh day after inoculation. The sections show a moderate
amount of connective tissue separating the islands of cartilage which
are several times the size of the islands in the transplantable tissue.
The epithelial cells have all disappeared and the stroma of the trans-
planted fragment is much less cellular than in the original tumor,
there are more connective tissue fibrils and the cells of the stroma
are more elongated than in the original tumor. In other words,
the tissue which surrounds the islands of cartilage is a reaction
product chiefly, although part of it is possibly derived from the old
Digitized by
Google
J. W. Jobling. 126
stroma. The cartilage appears normal, has not been invaded with
connective tissue, and a small number of the cells are in active
mitosis.
The second fragment examined was also obtained from Monkey
1, but was excised on the sixteenth day. Sections from this frag-
ment show that there has been no marked change in the masses of
cartilage, but the intervening connective tissue between the small
masses is much more cellular. Many of the cells in the connective
tissue are oval and resemble the stroma cells of the original tumor,
and the amount of fibrillated tissue is less than in the fragment
removed on the eleventh day. A few mitotic figures are still
present.
The next fragment examined was the one removed from Monkey
2. This fragment had been nine days in Monkey i and five days in
Monkey 2, or in the two monkeys for fourteen days. Micro-
scopically the fragment is seen to be encapsulated with fibrous tissue,
which in the center is quite hjraline. The cartilage masses are
degenerating, ( i ) through loss of nuclei chiefly in the center and by
degeneration of nuclei at the periphery, and (2) by a diffuse blue
(hematoxylin) staining in the center of some of the masses, probably
due to a calcareous deposit In this situation the cartilage cells are
shrunken, clear spaces exist in the matrix about them, and the nuclei
are deformed and pyknotic. In the periphery of some of the carti-
lage masses mitosis is still to be seen.
The next fragment examined was also from Monkey 2. This
fragment had been for nine days in Monkey i, and for twenty days
in Monkey 2, or twenty-nine days in both animals.
Microscopically the sections are composed almost entirely of
fibrous and adipose tissue, and acctunulations of small (lymphoid)
cells and numerous eosinophiles. Nothing remains of the tissue
transplanted.
Following the operation of March 23 the patient became pro-
gressively worse, the growth of the tumor being even more rapid
than before, and she died on May 4, 1908. At each dressing of the
operation wound a portion of the mass would adhere to the gauze
or could be easily detached with forceps, and with one such piece
another transplantation experiment was made.
Digitized by
Google
126 Transplantation Experiments in Macacus Rhesus,
Histologically the main difference between the material received
at this time and that at the preceding operation consists in this : the
growth now contains no cartilage and is composed of the cellular
stroma and masses of epithelial cells often appearing as acini.
Mitoses were numerous in both stroma and epithelium.
Monkey 4. This animal had not previously been inoculated. It received six
fragments subcutaneously.
4th day. Fragments cannot be felt.
8th day. Distinctly visible. Measure 2 to 3 mm. in diameter.
I2th day. Fragments disappeared.
Monkey 5. This was also a new animal. It received four fragments sub-
cutaneously, and 2 c.c. of an emulsion, made by teasing the tumor in sterile
ascitic fluid, was injected into the peritoneal cavity.
4th day. All fragments show apparent growth and measure 2 to 3 mm.
Sth day. Removed one fragment for histological study.
8th day. Fragments measure about 3 mm.
1 2th day. Barely perceptible.
14th day. Disappeared.
Monkey 2. This animal had previously been inoculated with fragments from
Monkey i. These fragments had shown a marked increase in size, but at the
time of reinoculation, on the fifteenth day, had already begun to be absorbed
and measured about 3 mm. The monkey was now reinoculated in four places
with new tumor material.
4th day. The fragments from first inoculation measure 2 mm. in diameter,
those from the recent inoculation measure 2 to 7 mm.
Sth day. One of the new fragments measuring 7 mm. was excised. A
portion was used to inoculate monkey 6 in six places, and the remainder was
saved for study. The fragment excised was rather sharply circumscribed and
firm in consistence. Complete disappearance of fragments.
8th day. New fragments measure 7 mm.
I2th day. Fragments measure 3 mm.
14th day. Barely perceptible.
20th day. Disappeared.
Monkey $- — ^This animal had also been previously inoculated with a fragment
excised from Monkey i. These fragments had shown a marked increase in
size, but at the time of reinoculation on the fifteenth day had undergone com-
plete absorption.
It was reinoculated in four places with the new material.
4th day. The fragments measure from 4 to 7 mm.
8th day. Measure 6 to 7 mm.
I2th day. Measure 3 mm.
14th day. Measure 3 mm.
20th day. Disappeared.
Monkey 6. — This was a new animal and was inoculated in six places with
the fragment excised on the fifth day from Monkey 2.
The animal died two days later from an intercurrent disease. The frag-
ments, which appeared pale, were saved for sections.
Digitized by
Google
J. W. Jobling. 127
Some of the tumor material used to inoculate the monkeys in this
second experiment was also used to inoculate five rats and five mice.
These animals were examined every second day for some time,
but at no examination was there any evidence of the fragments
having increased in size.
Microscopical examination of the fragments excised on the fifth
day from Monkey 5 shows that it consists of fibrous tissue enclos-
ing masses of small cells (lymphocytes) and some leucocytes. No
epithelial cells or cartilage are present. The increase in size of the
fragment was caused in this case by inflammatory reaction.
Examination of the fragment removed on the fifth day from
Monkey 2, which was used to inoculate Monkey 6, showed prac-
tically the same conditions noted above. The eosinophilic cells were
more numerous than in other instances. The fragments removed
from Monkey 6, which died two days after inoculation, were all
necrotic.
From the experiments recorded, it is obvious that for a time the
grafts increase in size under the skin of the monkey and do not
increase under the skin of mice or rats. The microscopical appear-
ance of the excised fragments indicates clearly what the causes
of the increase in size are. It was found, in the first series of
experiments, that the inflammatory reaction about the graft had
been comparatively slight, that the epithelial elements had not sur-
vived but had entirely disappeared, and that the cartilage had not
only remained alive for sixteen days, but still showed mitosis at the
end of that period and had increased several fold in size as com-
pared with the largest islands contained in the original tissue.
Examination of sections made from the tumor material used in the
second series of experiments shows that no cartilage was present
and in the masses removed from the site of inoculation the epithelial
cells and stroma of the original graft have been replaced by inflam-
matory tissue. As stated above, two series of experiments were
conducted. In the first series only one monkey was inoculated with
the nodules developing in this one, and two others were inoculated
in six places each.
In the second experiment, with fresh tumor material, two new
Digitized by
Google
128 Transplantation Experiments in Macacus Bhesus.
monkeys were inoculated, and at the same time two of the old
monkeys were inoculated.
Reference to the detailed description given for each monkey
shows that in the three monkeys of the first series of experiments
the measurements of fragments gave almost exactly similar figures
on the corresponding days, but the examination of the second series
in which new tumor material was used shows that the same rela-
tions do not exist.
These points are very well shown in the accompanying table.
First series
Second series.
Approximate
size of piece
injeaed.
Day.
Monkey
Monkey
Monkey
Monkey
Monkey
Monkey
Monkey
Monkey
z.
2.
3-
4-
5-
2.
3-
6.
4
0.5 mm.
2
2
2
4.5
5.5
5
3
7
2
3
2.5
8
3
3
7
7
9
2-5
II
3
4
4
I
3
3
14
3
3
I
3
i6
2
2
I8
I
I
20
?
There were four monkeys in the second series; two had
never been inoculated and two had been used in the first set of
experiments. Reference to the measurements obtained with the
new monkeys shows that there was progressive increase in
size of the fragments up to the eighth day, but that following this
there was a rapid retrogression leading to complete absorption. In
the two monkeys which were reinoculated, the increase in size was
about twice that noted for the fragments in the other two monkeys
of this series. The maximum size was reached on the eighth day,
following which there was a progressive diminution in size, though
in one monkey on the fourteenth day the fragment measured three
millimeters. Complete disappearance was first noted on the six-
teenth day. The differences noted in this series between the
monkeys which were inoculated for the first time and those which
were reinoculated are probably due to the greater reaction in the
subcutaneous tissues of the latter, corresponding perhaps to the reac-
Digitized by
Google
J. W. Jobling. 129
tion first reported by Arthus^ which occurs in the subcutaneous
tissues of rabbits following a third or fourth injection of horse
serum.
Although the attempts to implant the tumor fragments obtained
from a human subject upon monkeys failed, yet it can be said that
certain of the fragments increased considerably — ^three or four times
the size of the original fragments — in the subcutaneous tissues of
the monkey and the increase continued up to the sixteenth day after
the transplantation. During that period organic connection of the
transplanted tissue with the monkey had taken place by means of
a grantdation tissue which had surrounded and invaded the frag-
ments. It was found that all the teratomatous elements of the
tumor did not survive or grow, but that only one element, the
cartilage, was able to survive and increase in size. At the end of
eleven days the epithelial cells had entirely disappeared and no trace
of them could be found. It is possible that some of the cellular
stroma of the original tumor persisted and formed part of the
enveloping stroma of the cartilaginous masses. The oval cells of
the stroma of the original tumor are indistinguishable from the oval
cells of the transplanted fragments. That the cartilage of the
implanted fragments has remained alive is shown by the excellent
condition of its cells and the absence of leucoc)rtic invasion, and that
it has actually grown is indicated by the mitoses which are almost
as nxmierous in the specimen secured sixteen days after implanta-
tion as in the original tumor, and by the obvious increase in the
size of the cartilaginous masses. Thus it has been found that
cartilaginous tumor tissue from men can remain alive and pro-
liferating for sixteen days at least when implanted beneath the skin
of certain monkeys, while epithelial cells, undergoing active mitosis,
derived from the same tumor, quickly die and disappear.
The fact is worth noting that the results of the implantation of
teratoma of human origin have been wholly different and entirely
negative in certain lower animals — rats and mice — ^which are
further removed in respect to evolution from man than is the
monkey. I would, therefore, point out that it is not permissible to
conclude from one set of experiments on widely removed species
' Arthus, Compt, rend, Soc, de biol, 1903, v, 817.
Digitized by
Google
130 Transplantation Experiments in Macacus Bhesus.
that tumor implantations between heterologous species is absolutely
impossible.
Finally, I am of the opinion that the results which we have
obtained in the experiments on monkeys are not to be interpreted as
an actual partial success in transplanting tumors from man to the
monkey, but rather as an illustration, among the highest vertebrates,
of the existence of a certain kind of imperfect obstacle to tumor
growth between species, which has been expressed by Ehrlich* in
his conception of atreptic immunity. In other words, the tumor
cells introduced into the body of the foreign species are not at once
wholly destroyed by the fluids and cells of the host, but are gradu-
ally exhausted by the want of a special and peculiar nutriment
which the foreign host cannot supply, although a sufScient quantity
is carried along at the time of inoculation to enable the foreign
tumor cells to survive for a period. On the exhaustion of this
peculiar nutritious element, the transplanted cells undergo a form of
slow starvation retrogression.
• Ehrlich, P., Arb, a. d. k. Inst, f, exper. Therapie, 1906, No. i, 84.
Digitized by
Google
Digitized by
Google
CONTENTS OF MONOGRAPH No. 1
FLKXinER, Simon, and Jobli^q, J. W. Studies upon a Transplant-
able Rat Tumor. Plates I-XVI %
JoBUNG, J. W. The Biology of a Mixed Tumor of the Rat . 5j^
Mentbn, Maud L. Experiments on the Influence of Radium |
Bromide on a Carcinomatous Tumor of the Rat . . . 7»
< i
JoBUNO, J. W. SpontftneooB Tamore of the Moase. Platbs j
XVII-XXVIII . , 8li
JoBLiNG, J. W. Transplantation Experiments in Macacus rhesus i
with a Carcinomatous Teratoma from Man . . . 12(j
The object of the Monographs issued by the Rockefeller Institute']
for Medical Research is to provide a form of publication for scientifio;
papers which are so extensive or require such elaborate illustration asi
to render them unsuitable for current periodical issues.
Tlie Monographs will be published at irregular periods, determined,
by the available material in hand.
.1
It is proposed to secure as wide a circulation of the separate Mono-
graphs as is practicable, either through the distribution of complimen*',
tary copies or by sale, or in both ways, as the special conditions of eaclC
issue may determine.
Manuscripts, for which publication in the Monographs is de^red, *
should be sent to thi» Editor of the Jodhnal of Experimental Medi-|
CINE, Rockpfeller Institute for Medical Research, 66th Street and!
Avenue A, New York, N. Y.
I
'i
i
i
Digitized by
Google
Digitized by
Google
Digitized by
Google
Digitized by
Google
Digitized by
Google
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11